No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
The potential of digital surface models based on aerial images for automated vegetation mapping
International Journal of Remote Sensing
Abstract
Segmentation of vegetation patches was tested using canopy height models (CHMs)
representing the height difference between digital surface models (DSMs), generated
by matching digital aerial images from the Z/I Digital Mapping Camera, and a digital
elevation model (DEM) based on airborne laser scanner data. Three different combinations
of aerial images were used in the production of the CHMs to test the effect of
flight altitude and stereo overlap on segmentation accuracy. Segmentation results were
evaluated using the standard deviation of photo-interpreted tree height within segments,
as well as by visual comparison to existing maps. In addition, height percentiles
extracted from the CHMs were used to estimate tree heights. Tree height estimation at
the segment level yielded root mean square error (RMSE) values of 2.0 m, or 15.1%,
and an adjusted coefficient of determination (adjusted R2) of 0.94 when using a CHM
from images acquired at an altitude of 1200 m above ground level (agl) and with an
along-track stereo overlap of 80%. When a CHM based on images acquired at 4800 m
agl and an overlap of 60% was used, the corresponding results were an RMSE of
2.2 m, or 16.0%, and an adjusted R2 of 0.92. Tree height estimation at the plot level
was most accurate for densely forested plots dominated by coniferous tree species
(RMSE of 2.1 m, or 9.8%, and adjusted R2 of 0.88). It is shown that CHMs based on
aerial images acquired at 4800 m agl and with 60% along-track stereo overlap are
useful for the segmentation of vegetation and are at least as good as those based on
aerial images collected at a lower flight altitude or with greater overlap.
... Landscape level imagery acquisitions for the purposes of forest inventory related photogrammetric analyses have been proven capable and effective for providing structural and spectral forest inventory information (Bohlin et al., 2012;Granholm et al., 2015;Honkavaara et al., 2009;Nurminen et al., 2013;White et al., 2013b). Aerial imagery acquisitions are often updated on a regular basis by national or regional mapping entities (Straub et al., 2013;, further underwriting the costs of using these data in forest inventories, and making aerial images a dependable data source with temporal depth . ...
... To remedy the issue of poor DAP derived DTM quality, co-located ALS derived DTMs can be integrated into the DAP processing stream for point cloud normalization (Bohlin et al., 2012;Goodbody et al., 2018;Granholm et al., 2015;Nurminen et al., 2013;Vastaranta et al., 2013;White et al., 2013b). Moreover, structural metrics derived from DAP point clouds that use the same terrain information for normalization to heights above ground readily facilitate multitemporal comparisons, while improving the long-term value of ALS acquisitions. ...
... The accuracy of attribute predictions using DAP are of comparable or higher quality than traditionally acquired inventories, while providing greater detail about their spatial distribution (Bohlin et al., 2012;Goodbody et al., 2016;Granholm et al., 2015;Nurminen et al., 2013;Rahlf et al., 2017;White et al., 2013b). Although studies vary dramatically in their parameterization, they form a solid basis for the decision to use DAP data for updating EFIs, as well as for continued research into effective acquisition and processing standards. ...
In order to sustainably manage forest resources, a contemporary, dynamic, and consistent description of their state and extent must exist. As well, there is a need for reliable information on the change to the forested land base to support future policy development and to act informatively on new and emerging issues. Experimentation and technological innovation have spurred remote sensing research to better characterize and inventory forests globally. This dissertation examines how, and to what extent, digital aerial photogrammetry (DAP) and associated spatial products are capable of informing forest planning and management. Alongside innovation in DAP, unmanned aerial systems (UAS) are becoming viable management tools for acquiring stereo-imagery. Fast operationalization, cost-effectiveness, and their ability to acquire high spatial and temporal resolution data sets makes UAS a niche operational inventory tool. To assess the capacity of these technologies, forests of differing stages of structural development, including post-harvest regeneration, and mature managed forest landscapes, were examined. DAP data from these sites were analyzed to determine how, and to what extent, standard and novel inventory attributes can be accurately derived under specific image capture and data model circumstances. I advocate and provide evidence throughout this dissertation that DAP is a technology with ample potential for integration into enhanced forest inventory (EFI) frameworks. This work elaborates on where, and under what conditions, DAP data is successful and limited in characterizing forests with the ultimate goal of improving information for operational, tactical, and strategic decision-making.
... Research focused on the effectiveness of DAP in enhanced forest inventory frameworks has found that prediction accuracies for standard inventory attributes are similar to ALS, though DAP is fundamentally reliant on the ALS-derived terrain model (Bohlin et al., 2012;Granholm et al., 2015;White et al., 2013). Acquisition altitude and image resolution, along-and across-track image overlap, acquisition illumination, and meteorological conditions are all factors that may influence the quality of the image matching process and the derived DAP point clouds (Baltsavias, 1999). ...
... Balancing the trade-offs between acquisition parameter variations and costs, as well as accuracy of derived forest structure characterizations are crucial to forest management planning. Benchmarking studies that have examined the influence of acquisition parameters have been rare, and by necessity are often ad-hoc, cover small areas, and consider a limited range of forest conditions (Bohlin et al., 2012;Granholm et al., 2015;White et al., 2013). Moreover, many of these studies have been based on simulations rather than actual acquisitions and, as noted above, some parameters such as across-track overlap or resolution are difficult to simulate effectively. ...
Research has demonstrated the utility of digital aerial photogrammetry (DAP) for area-based predictions of forest inventory attributes. To date, studies have used DAP data acquired with a range of spatial resolutions and image overlaps. The systematic benchmarking of DAP acquisition parameters remains an outstanding research and operational gap for forest applications. While the impact of along-track overlap on point cloud metrics and area-based attribute estimates can be readily simulated, the impact of image resolution or across-track overlap requires purpose-acquired data. Moreover, although increases in along-track overlap are enabled by digital camera systems, costs for increasing across-track overlap can be substantial and may negate the cost-effectiveness of DAP for forest inventory. Hence, determining the impacts of varying acquisition parameters is of practical value for inventory programs. Researchers and practitioners have often assumed that more overlap will result in better DAP data, and that minimal overlaps often associated with historic airborne image campaigns are inadequate to support DAP processing. In our study, we found no marked difference among 15 and 20 cm spatial resolutions and overlap scenarios unless across-track overlap was reduced to 40%. Mean differences between DAP metrics and the ALS reference generally increased with decreasing overlap, and mean differences were larger for lower height percentiles (p10). Estimates of canopy height using the p90 metric varied by a root mean squared difference (RMSD) of approximately 5% between 15 cm and 20 cm datasets when along-track overlap was greater than 40%. Lower height percentiles were more strongly impacted by overlaps and resolution. Cover metrics varied by 2% RMSD across all overlap scenarios and resolutions. Comparisons between forest types (conifer, deciduous, mixed), terrain slope and aspect, and ALS-derived canopy cover were conducted to determine whether significant mean differences existed between DAP and the ALS reference. Although some significant differences were found by forest type and terrain variables, significant differences were most commonly associated with canopy cover. Based on the results reported herein, along and across-track overlaps ≥ 60% result in DAP metrics that were more similar to ALS. Increasing across-track overlap from 60% to 80% did not consistently improve the level of agreement between DAP metrics and ALS reference metrics. Conversely, DAP metrics generated using across-track overlaps <60% resulted in metrics with greater differences from the ALS reference, and a greater range of variability in DAP metric values. Image acquisitions for forest inventory must consider a broad range of factors and herein we have quantified that increasing along- or across-track overlap beyond 60% does not improve agreement with ALS area-based point cloud metrics commonly used to model forest inventory attributes. Likewise, overlap that is <60% does result in greater differences with ALS reference. Other applications beyond forest inventory may have different overlap requirements.
... Recent research on DSM generation using image matching methods revealed a great potential of DSM ap-plication in forestry, especially in forest inventory (e.g. [11][12][13][19][20][21][22][23][24][25]. In forestry application, DSM is usually used in combination with DTM derived from image point clouds or airborne laser scanning (ALS) data. ...
... measurements from sample plots) following the example of recent worldwide studies (e.g. [11][12][13][19][20][21][22][23][24][25]. ...
Background and Purpose: Recent research on generation of digital surface models (DSMs) using image matching methods revealed a great potential of DSM application in forestry, especially in forest inventory. However, research dealing with DSM generation from digital aerial images are still lacking in Croatia. Therefore, the main objective of this study was to present the workflow for generating high density DSM from colour infrared (CIR) digital stereo aerial images using area-based image matching algorithm.
Materials and Methods: The high density DSM was generated from colour infrared digital aerial stereo images using Dense DTM algorithm of PHOTOMOD software - an area-based image matching algorithm which operates on the principle of cross-correlation approach. To evaluate the quality of the generated DSM, an agreement assessment with manual stereo measurements was conducted over three different land cover classes (forests, shrubs, grasslands) using the same images as for DSM generation.
Results: The good vertical agreement between the generated DSM and stereo measurement was achieved for all three land cover classes present at the research area. The highest vertical agreement was obtained for the grassland land cover class (RMSE=0.36), slightly lower for forest (RMSE=0.62), whereas the lowest vertical agreement was obtained for shrub land cover class (RMSE=0.83).
Conclusions: The results of this research are very promising and suggest that the high density DSM generated from digital aerial stereo images and by using the proposed methodology has the potential to be used in forestry, primarily in forest inventory. Therefore, further research should be focused on generation of CHM by subtracting available DTM from the high density DSM and on the examination of its potential for deriving various forest attributes.
... Among all remote sensing data, over the last two decades, airborne LiDAR data were the dominant type employed for individual tree attributes estimation since they provide rapid and accurate information with adequate spatial detail about forest variables (Lim et al., 2003). Unmanned Aerial Vehicles (UAV) that acquire optical data with very high spatial resolutions have emerged as a feasible and costeffective alternative for forestry as a result of current technological advancements in sensors and digital image processing systems (Getzin et al., 2014;Granholm et al., 2015;Zhang et al., 2016). Notably, the development of the Structure from Motion (SfM) technique allows the extraction of fast and reliable photogrammetric point cloud data from UAV-based optical data through image matching approaches (Díaz-Varela et al., 2015;Zarco-Tejada et al., 2014). ...
In this study, a graph-based approach was developed for detecting individual trees in a broadleaf, complex forest region based on UAV-derived point cloud data. Horizontal cross-sections at different heights were applied to the Canopy Height Model (CHM) to extract initial candidates for graph nodes. The graph was processed in multiple steps, and individual treetop locations were detected based on graph nodes’ properties. The impact of various parameters, such as minimum area of connected components and minimum tree heights, on the performance of the developed method was investigated. The evaluation step demonstrated the potential of the proposed graph-based method for individual tree detection in a complex forest region in Mazandaran, Iran.
The source code of the proposed algorithm can be found at https://github.com/Seyed-Ali-Ahmadi/Graph-based_ITCD.
... Additionally, there was no clear bias in regard to tree plots from specific sites, thus showing that the observed trends do not clearly change between our study sites. Our results comparing DAP DSMs to plot tree heights are in agreement with other studies (Granholm et al., 2015;Mielcarek et al., 2020;Straub and Stepper, 2016). Additionally, the two study sites have relatively even distribution across the trendline showing that the model is adequate for both study sites. ...
In the absence of complete lidar coverage, digital surface models (DSMs) and point clouds produced from the United States Department of Agriculture National Agriculture Imagery Program (NAIP) are increasingly being analyzed for quality and application feasibility. This study compared canopy heights derived from NAIP DSMs (10 m) and point clouds to those derived from lidar data collected over Mountain Lake Biological Station and the Great Smoky Mountains Twin Creeks Site by the National Ecological Observatory Network (NEON) Airborne Observation Platform for 62 mixed deciduous tree plots. Mean dominant height (MDH) was estimated using lidar and the NAIP products using the 90th percentile of heights in a given plot as the independent variable for both the lidar- and NAIP-derived point clouds. The dependent variable was field-measured MDH, calculated using the four tallest trees for each 0.04-hectare plot based on the NEON woody vegetation structure dataset. All data (field and remotely sensed) were collected in 2018. Using maximum likelihood spatial error model for all analyses, the NAIP DSM (10 m resolution) resulted in a strong relationship with MDH (coefficient of determination (R2) = 0.90, standard error (SE) = 1.71 m). However, the 90th percentiles of heights derived from the point clouds were better at estimating MDH than was the comparatively coarse resolution DSM (NAIP point clouds: R2 = 0.94, SE = 1.40 m; lidar: R2 = 0.95, SE= 1.29 m, respectively) and are strongly correlated to each other (R2 = 0.99, SE = 0.68 m). The main limitation of the NAIP datasets was found to be where shadowing occurred due to steep terrain in the Great Smoky Mountain site. These areas resulted in erroneously high vegetation heights. Mean dominant heights estimated using NAIP DSMs and point clouds are thus comparable to those estimated using lidar data in these closed-canopy temperate deciduous forests where shadowing from steep terrain is not present. The utility of both the NAIP-derived 10 m DSM and the point clouds for estimating tree heights paves the way for statewide mapping of heights over the deciduous forests in Tennessee, Virginia, and possibly beyond.
... Active remote sensing sensors and structure-from-motion techniques utilized in aerial image processing produce three-dimensional models of the target scene, which allows for structural representations of vegetation [59,61,76]. The use of UAVs is expanding for mapping forest canopy structure, e.g., [60,76], and more recently, these techniques were applied to other vegetation types, e.g., [51,68,77,78]. In rangeland environments, shrub and grass structures may help in differentiating vegetation types [68]; therefore, we produced a canopy height model for our study sites. ...
Grasslands and shrublands exhibit pronounced spatial and temporal variability in structure and function with differences in phenology that can be difficult to observe. Unpiloted aerial vehicles (UAVs) can measure vegetation spectral patterns relatively cheaply and repeatably at fine spatial resolution. We tested the ability of UAVs to measure phenological variability within vegetation functional groups and to improve classification accuracy at two sites in Montana, U.S.A. We tested four flight frequencies during the growing season. Classification accuracy based on reference data increased by 5–10% between a single flight and scenarios including all conducted flights. Accuracy increased from 50.6 to 61.4% at the drier site, while at the more mesic/densely vegetated site, we found an increase of 59.0 to 64.4% between a single and multiple flights over the growing season. Peak green-up varied by 2–4 weeks within the scenes, and sparse vegetation classes had only a short detectable window of active phtosynthesis; therefore, a single flight could not capture all vegetation that was active across the growing season. The multi-temporal analyses identified differences in the seasonal timing of green-up and senescence within herbaceous and sagebrush classes. Multiple UAV measurements can identify the fine-scale phenological variability in complex mixed grass/shrub vegetation.
... Recent advances in the fields of Unmanned Aerial Vehicle (UAV) technique and image processing have widened the scope of forestry remote sensing, and there is potential for conducting highly accurate forest inventory economically and effectively (Zhang et al. 2016;Mohan et al. 2017). Low consumer-grade cameras attached to UAV are increasing rapidly used by forest managers to record general forest situations at small scales for low cost and high manoeuvrability (White et al. 2013;Getzin, Nuske, and Wiegand 2014;Granholm et al. 2015;Zhang et al. 2016). How to retrieve more information from these UAV imageries is still in progress. ...
Low consumer-grade cameras attached to small unmanned aerial vehicles (UAV) can easily acquire high spatial resolution images, leading to convenient forest monitoring at small-scales for forest managers. However, most studies were carried out in the low canopy density and flat ground plantations to detect individual trees. We selected overlapping canopy plantation in mountainous area in the eastern of China and acquired high spatial resolution UAV RGB images to detect individual trees. A total of 402 reference trees were located in three rectangle plots (900 m²). To enhance the confidence of the tested individual tree detection method, clear-cutting and Real-Time Kinematic (RTK) were used to obtain the truth values in the plots. A novel method for semi-automatic individual tree detection was proposed based on a local-maximum algorithm and UAV-derived DSM data (LAD) in this study. The detection accuracy of LAD was compared with commonly used methods based on UAV-derived orthophoto images, local-maximum algorithm (LAO), object-oriented feature segmentation (OFS), multiscale segmentation technique (MST) and manual visual interpretation (MVI). The overall accuracy (OA (%) decreased in the order of LAD (84.5%) > MST (69.1%) > OFS (65.1%) > MVI (64.1%) > LAO (59.1%). LAD had only 15.5%s omission errors (OM (%), which was less than half of the other four methods in comparison. It was noteworthy that MVI had 35.9% OM %, which revealed that MVI should be used carefully as the truth value. LAD showed similar repeated detection error (RP (%) and completely wrong detection (CW (%), while the other four methods had obviously higher CW % than the RP %. From our results, it can be concluded that the proposed LAD method may help improving the accuracy of individual tree detection to an acceptable accuracy (>80%) in dense mountain forests, and has practical advantages in future research direction to assess tree attributes from UAV RGB image.
... Ota et al. (2015) estimated aboveground biomass in seasonal tropical forest by using point clouds derived from digital airborne imagery. Granholm et al. (2015) used digital surface models based on aerial images for automated vegetation mapping. Aerial images cover a large extent, and are more appropriate for generating digital surface model over large areas. ...
Characterizing vegetation structure is an important component for understanding ecological recovery on non-permanent human footprint features in forests. However, current approaches to measuring vegetation structure rely on field protocols that are costly and difficult to scale. Compared to traditional field methods, UAV (unmanned aerial vehicle) photogrammetry has shown great promise in characterizing vegetation structure in a more cost-efficient way. In this research, I used a point-intercept sampling strategy to conduct a comparison of UAV-based estimates and field measurements at two scales: (i) point level and (ii) site (plot) level. I found that at the aggregated site level, UAV photogrammetry alone could replace traditional field-based vegetation surveys of mean vegetation height across the range of conditions assessed in this study, though significant differences remain between remote-and field-based vegetation surveys at point level. Cost analysis indicates that using UAV point clouds alone provides substantial cost-saving over traditional field vegetation surveys.
... Similarly, forest-growing stock experienced the same trend with a time lag, and even doubled during the last 50 years [3]. analysis compared to ALS or InSAR data [45,50,51]. Furthermore, multi-temporal aerial photographs also facilitate estimation of growth increment or canopy change detection that may improve the precision of MSNFI estimates. ...
Multisource forest inventory methods were developed to improve the precision of national forest inventory estimates. These methods rely on the combination of inventory data and auxiliary information correlated with forest attributes of interest. As these methods have been predominantly tested over coniferous forests, the present study used this approach for heterogeneous and complex deciduous forests in the center of France. The auxiliary data considered included a forest type map, Landsat 8 spectral bands and derived vegetation indexes, and 3D variables derived from photogrammetric canopy height models. On a subset area, changes in canopy height estimated from two successive photogrammetric models were also used. A model-assisted inference framework, using a k nearest-neighbors approach, was used to predict 11 field inventory variables simultaneously. The results showed that among the auxiliary variables tested, 3D metrics improved the precision of dendrometric estimates more than other auxiliary variables. Relative efficiencies (RE) varying from 2.15 for volume to 1.04 for stand density were obtained using all auxiliary variables. Canopy height changes also increased RE from 3% to 26%. Our results confirmed the importance of 3D metrics as auxiliary variables and demonstrated the value of canopy change variables for increasing the precision of estimates of forest structural attributes such as density and quadratic mean diameter.
... Regarding image space error, an average re-projection error of 0.48 pixels (maximum value of 0.91 pixels) was achieved. In any case, there is a growing need to explore the impacts of different image acquisition and processing parameters on DAP SfM derived outputs such as flying height, software package choice and parameter settings [60,62,63]. ...
Remote sensing is revolutionizing the way in which forests studies are conducted, and recent technological advances, such as Structure from Motion (SfM) photogrammetry from Unmanned Aerial Vehicle (UAV), are providing more efficient methods to assist in REDD (Reducing Emissions from Deforestation and forest Degradation) monitoring and forest sustainable management. The aim of this work was to develop and test a methodology based on SfM from UAV to generate high quality Digital Terrain Models (DTMs) on teak plantations (Tectona grandis Linn. F.) situated in the Coastal Region of Ecuador (dry tropical forest). UAV overlapping images were collected using a DJI Phantom 4 Advanced© quadcopter during the dry season (leaf-off phenological stage) over 58 teak square plots of 36 m side belonging to three different plantations located in the province of Guayas (Ecuador). A workflow consisting of SfM absolute image alignment based on field surveyed ground control points, very dense point cloud generation, ground points filtering and outlier removal, and DTM interpolation from labeled ground points, was accomplished. A very accurate Terrestrial Laser Scanning (TLS) derived ground points were employed as ground reference to estimate the UAV-SfM DTM vertical error in each reference plot. The plot-level obtained DTMs presented low vertical bias and random error (−3.1 cm and 11.9 cm on average, respectively), showing statistically significant greater error in those reference plots with basal area and estimated vegetation coverage above 15 m2/ha and 60%, respectively. To the best of the authors’ knowledge, this is the first study aimed at monitoring of teak plantations located in dry tropical forests from UAV images. It provides valuable information that recommends carrying out the UAV image capture during the leaf-off season to obtain UAV-SfM derived DTMs suitable to serve as ground reference in supporting teak plantations inventories.
... The main benefit of the photogrammetric canopy model is that no separate flights are required for the acquisition of 3D data and imagery, which often have different flight parameters in relation to the imaging altitude and coverage in case of ALS and 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 aerial imagery. Furthermore, aerial images are typically acquired at frequent intervals for purposes other than forestry such as mapping and surveying, which increases their availability compared to ALS data (Granholm et al. 2015;Stepper et al. 2016). Therefore, aerial images have significant potential for operational use in forestry applications (Bohlin et al. 2012). ...
Optical 2D remote sensing techniques such as aerial photographing and satellite imaging have been used in forest inventory for a long time. During the last 15 years, airborne laser scanning (ALS) has been adopted in many countries for the estimation of forest attributes at stand and sub-stand levels. Compared to optical remote sensing data sources, ALS data are particularly well-suited for the estimation of forest attributes related to the physical dimensions of trees due to its 3D information. Similar to ALS, it is possible to derive a 3D forest canopy model based on aerial imagery using digital aerial photogrammetry. In this study, we compared the accuracy and spatial characteristics of 2D satellite and aerial imagery as well as 3D ALS and photogrammetric remote sensing data in the estimation of forest inventory variables using k-NN imputation and 2469 National Forest Inventory (NFI) sample plots in a study area covering approximately 5800 km². Both 2D data were very close to each other in terms of accuracy, as were both the 3D materials. On the other hand, the difference between the 2D and 3D materials was very clear. The 3D data produce a map where the hotspots of volume, for instance, are much clearer than with 2D remote sensing imagery. The spatial correlation in the map produced with 2D data shows a lower shortrange correlation, but the correlations approach the same level after 200 meters. The difference may be of importance, for instance, when analyzing the efficiency of different sampling designs and when estimating harvesting potential. © 2017, Finnish Society of Forest Science. All rights reserved.
... A number of recent studies emphasized the great potential of dense point clouds and DSMs derived by image matching of digital aerial stereo images in combination with accurate DTMs for the CHMs and forest attributes estimation [3,19,22,[24][25][26][27]. Moreover, comparison studies which evaluated different remote sensing data considered this image-based approach as a cost-effective alternative to ALS in forest inventory applications [10][11][12][28][29][30][31][32][33][34]. ...
Digital aerial photogrammetry has recently attracted great attention in forest inventory studies, particularly in countries where airborne laser scanning (ALS) technology is not available. Further research, however, is required to prove its practical applicability in deriving three-dimensional (3D) point clouds and canopy surface and height models (CSMs and CHMs, respectively) over different forest types. The primary aim of this study is to investigate the applicability of image-based CHMs at different spatial resolutions (1 m, 2 m, 5 m) for use in stand-level forest inventory, with a special focus on estimation of stand-level merchantable volume of even-aged pedunculate oak (Quercus robur L.) forests. CHMs are generated by subtracting digital terrain models (DTMs), derived from the national digital terrain database, from corresponding digital surface models (DSMs), derived by the process of image matching of digital aerial images. Two types of stand-level volume regression models are developed for each CHM resolution. The first model is based solely on stand-level CHM metrics, whereas in the second model, easily obtainable variables from forest management databases are included in addition to CHM metrics. The estimation accuracies of the stand volume estimates based on stand-level metrics (relative root mean square error RMSE% = 12.53%–13.28%) are similar or slightly higher than those obtained from previous studies in which stand volume estimates were based on plot-level metrics. The inclusion of stand age as an independent variable in addition to CHM metrics improves the accuracy of the stand volume estimates. Improvements are notable for young and middle-aged stands, and negligible for mature and old stands. Results show that CHMs at the three different resolutions are capable of providing reasonably accurate volume estimates at the stand level.
... The majority of research on using height data in combination with spectral data for the classification of vegetation has been primarily made related to forestry (e.g., [9][10][11]). However, there are considerable differences in height and spatial extent between trees and aquatic plants. ...
Monitoring of aquatic vegetation is an important component in the assessment of freshwater ecosystems. Remote sensing with unmanned aircraft systems (UASs) can provide sub-decimetre-resolution aerial images and is a useful tool for detailed vegetation mapping. In a previous study, non-submerged aquatic vegetation was successfully mapped using automated classification of spectral and textural features from a true-colour UAS-orthoimage with 5-cm pixels. In the present study, height data from a digital surface model (DSM) created from overlapping UAS-images has been incorporated together with the spectral and textural features from the UAS-orthoimage to test if classification accuracy can be improved further. We studied two levels of thematic detail: (a) Growth forms including the classes of water, nymphaeid, and helophyte; and (b) dominant taxa including seven vegetation classes. We hypothesized that the incorporation of height data together with spectral and textural features would increase classification accuracy as compared to using spectral and textural features alone, at both levels of thematic detail. We tested our hypothesis at five test sites (100 m × 100 m each) with varying vegetation complexity and image quality using automated object-based image analysis in combination with Random Forest classification. Overall accuracy at each of the five test sites ranged from 78% to 87% at the growth-form level and from 66% to 85% at the dominant-taxon level. In comparison to using spectral and textural features alone, the inclusion of height data increased the overall accuracy significantly by 4%-21% for growth-forms and 3%-30% for dominant taxa. The biggest improvement gained by adding height data was observed at the test site with the most complex vegetation. Height data derived from UAS-images has a large potential to efficiently increase the accuracy of automated classification of non-submerged aquatic vegetation, indicating good possibilities for operative mapping.
... Further, while modern digital airborne cameras greatly facilitate the acquisition of multiple overlapping images (Bohlin et al. 2012;Straub et al. 2013), and while there has been some work to evaluate the influence of image overlap (Ofner et al. 2006;Granholm et al. 2015), there are as yet no comprehensive guidelines to ensure optimal flight planning for modern SfM-MVS aerial photography in support of forest mapping. ...
Airborne LiDAR data is now commonly acquired by the Australian plantation sector in order to generate accurate digital terrain models and canopy height models at high spatial resolution for resource assessment estimates. However, these airborne surveys are relatively expensive and there is a desire to identify more affordable options for collecting or updating this information. This review presents alternative approaches to deriving canopy height models, including the use of stereo optical imagery from satellites, manned and unmanned airborne platforms, and the use of synthetic aperture radar. In addition, we illustrate the potential of airborne photogrammetry with multi-view dense point matching to produce an accurate, hybrid photo-LiDAR canopy height model at high spatial resolution along a transect covering stands of several softwood tree species.
... One of the current issues associated with DAP is a lack of standards or best practices surrounding appropriate image inputs for point-cloud generation (White, Stepper, et al. 2015). Although optimal specifications have been suggested (Leberl et al. 2010), these likely vary by application and stand conditions, and, to date, there has been little benchmarking done for forest targets specifically (Haala et al. 2010, Remondino et al. 2014); however, more recent studies have begun to explore the impacts of acquisition parameters on information outcomes (e.g., Granholm et al. 2015;Ota et al. 2015). Because a variety of image parameters, including ground-sampling distance (GSD) and image overlap are being used in research contexts, comparison of inventory outcomes is challenging. ...
Forest inventory and management requirements are changing rapidly in the context of an increasingly complex set of economic, environmental, and social policy objectives. Advanced remote sensing technologies provide data to assist in addressing these escalating information needs and to support the subsequent development and parameterization of models for an even broader range of information needs. This special issue contains papers that use a variety of remote sensing technologies to derive forest inventory or inventory-related information. Herein, we review the potential of 4 advanced remote sensing technologies, which we posit as having the greatest potential to influence forest inventories designed to characterize forest resource information for strategic, tactical, and operational planning: airborne laser scanning (ALS), terrestrial laser scanning (TLS), digital aerial photogrammetry (DAP), and high spatial resolution (HSR)/very high spatial resolution (VHSR) satellite optical imagery. ALS, in particular, has proven to be a transformative technology, offering forest inventories the required spatial detail and accuracy across large areas and a diverse range of forest types. The coupling of DAP with ALS technologies will likely have the greatest impact on forest inventory practices in the next decade, providing capacity for a broader suite of attributes, as well as for monitoring growth over time.
... LSB-Snakes could be considered as an extension of the standard MPGC algorithm, in which parameters of linear B-spline functions in object space were directly estimated to refine edge matching. 18 In the past 10 years, this method was widely used in automated vegetation mapping, 19 urban road matching, 20 and many other fields. Although this linear LSB-Snakes method can successfully match more than 80% of the edges computed from the multiple primitive multiimage matching procedure, some limitations of this method were also discovered. ...
This paper presents a point cloud optimization method of low-altitude remote sensing image based on least square matching (LSM). The proposed method is designed to be especially effective for addressing the conundrum of stereo matching on the discontinuity of architectural structures. To overcome the error matching and blur on building discontinuities in three-dimensional (3-D) reconstruction, a pair of mutually perpendicular patches is set up for every point of object discontinuities instead of a single patch. Then an error equation is built to compute the optimal point according to the LSM method, space geometry relationship, and collinear equation constraint. Compared with the traditional patch-based LSM method, the proposed method can achieve higher accuracy 3-D point cloud data and sharpen the edge. This is because a geometric mean patch in patch-based LSM is the local tangent plane of an object's surface. Using a pair of mutually perpendicular patches instead of a single patch evades the problem that the local tangent plane on the discontinuity of a building did not exist and highlights the edges of buildings. Comparison studies and experimental results prove the high accuracy of the proposed algorithm in low-altitude remote sensing image point cloud optimization. © 2016 Society of Photo-Optical Instrumentation Engineers (SPIE).
Inventorying forest ecosystems is an essential part of forest management planning. However, it is quite costly and time-consuming, particularly for larger areas. Recently, significant developments have been made in unmanned aerial vehicle (UAV) technology to improve the cost and time efficiency in forest inventory. Therefore, UAV images have become one of the inventory tools that produces data with high spatial resolution in determining forest resources. This study aims to investigate the contribution of UAV data to forest inventory in a case study area with a total of 30 sample plots located in pure and natural Crimean pine (Pinus nigra J.F. Arnold ssp. pallasiana (Lamb.) Holmboe) stands in the Black Sea backward region of Türkiye. Total tree height (h) and stem volume (v) were recorded at individual tree level (n = 367), and the number of trees (N), mean height (hmean), top height (htop), stand basal area (BA) and stand volume (V) were calculated at sample plot level (n = 30) from both the field and UAV-based data. Pearson’s correlation coefficients (r) for h and v were 0.96 and 0.72, respectively, the highest correlation at the sample plot level was observed for the hmean - htop (r = 0.96), while the lowest correlation was found for BA (r = 0.54). The suitability of the observation and prediction values was assessed using a t-test at both individual tree and sample plot levels. According to the t-test results, the observation and prediction values for h, v, hmean, htop, BA and V metrics were found to be compatible (p > 0.05), but not for N (p < 0.05). Overall results indicated that UAV technology has a potential to be used in forest inventory and can contribute to the determination of individual tree and stand metrics. Thereby, it saves cost and time in forest inventory studies and helps monitoring the dynamic structure of the forest ecosystem with an effective approach in forest inventory.
This research introduces a scientific methodology to identify areas affected by gully erosion using Geographic Object Based Image Analysis (GEOBIA) and Random Forest (RF) supervised machine learning. The GEOBIA and RF were applied in Besni district, which has a Mediterranean climate, of Adiyaman province in Turkey by including many factors in the model. Estimation Scale Parameter (ESPII) algorithm was used in the segmentation phase. The novelty of this study is the implementation of RF supervised classification algorithm to classify a large number of objects determined after the segmentation process, due to the large size of the study area. Therefore, open access data has been evaluated with high classification accuracy without the need for labor. Precision, Recall and F1-Score values were calculated using true positive (TP), true negative (TN), false positive (FP) and false negative (FN) values based on field observations and Google Earth images of the study area. The TP, TN, FP and FN values were 0.90, 0.95 and 0.92, respectively. In addition, a Kappa-index was calculated as 0.88. The gully erosion map obtained using aforementioned methodology can be used to take necessary measures to prevent further degradation and plan sustainable land uses.
The paper considers the methods of acquisition and processing of optical data from small Unmanned Air Vehicles (UAVs) ― photogrammetric point clouds and derivative 3d-models — for the automatic extraction of explicit structure variables in sparse boreal forests of the central part of Kola Peninsula.
We review main technological issues of UAV optical surveys, present flowcharts of point clouds classification for the extraction of canopy height model (CHM), further CHM analysis, tree-level and areabased
estimation of structural forest variables. Main tree-level variables are crown heights and extent; for forest stands CHM analysis leads to gridded data on tree canopy heights, amount of canopy peaks and tree density, share of tree cover. The definite limitations of optical photogrammetry connected with CHM extraction in dense forests can be partly overcome due to the complex use of point clouds from summer and winter (leaf-off) surveys and independent processing flow of CHM in forest stands with sparser and denser tree canopies.
Purpose of Review
Three-dimensional (3D) data on forest structure have transformed the level of detail and accuracy of forest information. While these 3D data have primarily been derived from airborne laser scanning (ALS), there has been growing interest in the use of 3D data derived from digital aerial photogrammetry (DAP) and image-matching algorithms. In particular, research and operational forestry communities are interested in using DAP data to update existing ALS-derived enhanced forest inventories. Although DAP depends on accurate terrain information provided by ALS to normalize digital surface models to heights above ground, in an inventory update scenario, DAP data currently have cost advantages over repeat ALS acquisitions.
Recent Findings
Extensive research across a broad range of forest types has demonstrated that DAP data can provide comparable accuracies to ALS for estimating inventory attributes such as volume, basal area, and height when used in an area-based approach with co-located ground plot information.
Summary
Herein, we review research relevant to the use of DAP for updating area-based forest inventories in subsequent inventory cycles, highlighting issues and opportunities for DAP data in this context. We examine the use of DAP for area-based forest inventory applications, comparing data inputs, algorithms, and outcomes across numerous studies and forest environments. Lastly, we outline outstanding research gaps that require further inquiry including benchmarking of acquisition parameters and image-matching algorithms.
The National Inventory of Landscapes in Sweden (NILS) was established by the Swedish Environmental Protection Agency to provide data for policy-makers in the country. Its main role is to determine the status of (and changes in) the Swedish landscape, either as a consequence of natural and/or anthropogenic disturbances, or because of ecological processes. In 2017, data from NILS will become available on the NILS data portal for analysis by researchers and other interested parties, such as governmental bodies. NILS’s data collection covers all terrestrial areas and identifies variables such as land cover, land use (including historical land use), tree, shrub, field and ground vegetation and surfaces. The program consists of two parallel inventories, a field inventory and a remote sensing inventory, with both covering the nation in five-year rotations. The field inventory employs a large group of field-workers, and all sample squares are also inventoried using aerial near-infrared stereo imagery provided by the Swedish Land Survey. The first rotations of field data will soon be available, and the data from the first rotation of the remote sensing component (2003–2007) is already available. As part of the EU, Sweden is currently updating the mapping of land cover data, using the national coverage of airborne Light Detection and Ranging (LIDAR) and the European satellites of the Sentinel series. While at the same time strengthening the statistical estimates made from the NILS data. The principal role of the NILS program is to provide the reference data for both the classification of chosen vegetation types and for the validation of the results. The data collection for this mapping will start in 2017, using remote sensing and modelling, and sampling additional inventory areas for reference data.
This study had the aim of investigating the utility of image-based point cloud data for estimation of vertical canopy cover (VCC). An accurate measure of VCC based on photogrammetric matching of aerial images would aid in vegetation mapping, especially in areas where aerial imagery is acquired regularly. The test area is located in southern Sweden and was divided into four vegetation types with sparse to dense tree cover: unmanaged coniferous forest; pasture areas with deciduous tree cover; wetland; and managed coniferous forest. Aerial imagery with a ground sample distance of 0.24 m was photogrammetrically matched to produce dense image-based point cloud data. Two different image matching software solutions were used and compared: MATCH-T DSM by Trimble and SURE by nFrames. The image-based point clouds were normalized using a digital terrain model derived from airborne laser scanner (ALS) data. The canopy cover metric vegetation ratio was derived from the image-based point clouds, as well as from raster-based canopy height models (CHMs) derived from the point clouds. Regression analysis was applied with vegetation ratio derived from near nadir ALS data as the dependent variable and metrics derived from image-based point cloud data as the independent variables. Among the different vegetation types, vegetation ratio derived from the image-based point cloud data generated by using MATCH-T resulted in relative root mean square errors (rRMSE) of VCC ranging from 6.1% to 29.3%. Vegetation ratio based on point clouds from SURE resulted in rRMSEs ranging from 7.3% to 37.9%. Use of the vegetation ratio based on CHMs generated from the image-based point clouds resulted in similar, yet slightly higher values of rRMSE.
In recent years, digital aerial photogrammetry has emerged as an alternative method to airborne laser scanning in three-dimensional modelling of forest areas, especially for the generation of digital surface models (DSMs). In forest inventory, DSM is usually used in combination with the corresponding digital terrain model for the generation of canopy height model (CHM), which is then used as a basis for deriving various tree and forest attributes.
The main aim of this research was to examine the vertical accuracy of DSMs of different spatial resolutions over the forest area, and with new findings contribute to the application of digital photogrammetry in forest science and practice. For that purpose, DSMs with spatial resolution (pixel size) of 0.3 m (DSM0.3), 0.5 m (DSM0.5), 1 m (DSM1), 2 m (DSM2) and 5 m (DSM5) were generated by image matching of digital aerial images for the area of lowland pedunculate oak forests (management unit Kunjevci, Forest Administration Vinkovci). The vertical accuracy of DSMs was evaluated by comparing manually stereo measured elevations of 294 tree tops with the elevations of planimetrically corresponding DSMs points.
As expected, the highest accuracy was obtained for DSM0.3 (root mean square error, RMSE = 0.76 m; mean error, ME = –0.03 m). Almost equal accuracy was obtained for DSM0.5 (RMSE = 0.76 m; ME = –0.05 m) and DSM1 (RMSE = 0.76 m; ME = –0.07 m), slightly lower for DSM2 (RMSE = 0.84 m; ME = –0.16 m), whereas the lowest accuracy was obtained for DSM5 (RMSE = 1.31 m; ME = –0.54 m).
The accuracy comparison showed that the decreasing of spatial resolution (pixel size) of raster based DSMs from 0.3 m to 1 m, does not significantly affect their vertical accuracy. With further decreasing of spatial resolution to 2 m, and especially to 5 m, the vertical accuracy of DSMs also decreases. In the light of the obtained results, further studies should be focused on research of possibilities of application of DSMs of different spatial resolution in forest inventory, namely: DSM0.3 and DSM0.5 for obtaining information at tree level, DSM0.5 and DSM1 at plot level, and DSM1, DSM2 and DSM5 at stand level.
For decades, aerial photo interpretation has been, and to a good extent is still, the method of choice for producing fine-scale native forest stand mapping. Recent computer techniques have eased the task of the interpreter, who is now able to delineate polygons through on-screen digitising in a geographical information system (GIS) environment. Even with these advances, a great deal of skill is required in the polygon delineation. In an effort to contribute to the automation of this process, we introduce an open-source object-based solution to the mapping of forest stand boundaries using attributes derived from digital aerial photography and laser scanning data acquired over a study area in the Victorian Central Highlands. This methodology transforms remotely sensed imagery (single or multichannel) and canopy raster layers derived from laser scanning (lidar) into polygon vector layers. It is intended that the resultant polygon layer should resemble the product derived by an aerial interpreter, without any prior knowledge of the scene. The derived product aims to produce a layer comprised of relatively homogeneous polygons all exceeding a minimum size. The derived product is meant to be a preliminary template aimed at reducing time and effort in manual digitisation. The relationship between spectral, texture and laser scanning derived features for forest stand boundary delineation and human interpreted boundaries is not straight forward. The interpreter however, can aggregate and sometimes correct the automated delineated regions by simple drag-and-click operations This approach is relatively cheap and flexible, being a workable compromise between fully automated image interpretation which requires further research for acceptable levels of accuracy and reliability, and manual segmentation and classification. Preliminary results are encouraging, both in regard to automating the process and the delivery of robust delineation of stand boundaries in native forest landscapes. Future research will focus on appropriate input resolution to reduce computation requirements and improved data fusion methods to obtain more accurate forest stand delineation.
Airborne laser scanning (ALS) has demonstrated utility for forestry applications and has renewed interest in
other forms of remotely sensed data, especially those that capture three-dimensional (3-D) forest characteristics. One such data source results from the advanced processing of high spatial resolution digital stereo imagery (DSI) to generate 3-D point clouds. From the derived point cloud, a digital surface model and forest vertical information with similarities to ALS can be generated. A key consideration is that when developing forestry related products such as a canopy height model (CHM), a high spatial resolution digital terrain model (DTM), typically from ALS, is required to normalize DSI elevations to heights above ground. In this paper we report on our investigations into the use of DSI-derived vertical information for capturing variations in forest structure and compare these results to those acquired using ALS. An ALS-derived DTM was used to provide the spatially detailed ground surface elevations to normalize DSI-derived heights. Similar metrics were calculated from the vertical information provided by both DSI and ALS. Comparisons revealed that ALS metrics provided a more detailed characterization of the canopy surface including canopy openings. Both DSI and ALS metrics had similar levels of correlation with forest structural attributes (e.g., height, volume, and biomass). DSI-based models predicted height, diameter, basal area, stem volume, and biomass with root mean square (RMS) accuracies of 11.2%, 21.7%, 23.6%, 24.5%, and 23.7%, respectively. The respective accuracies for the ALS-based predictions were 7.8%, 19.1%, 17.8%, 17.9%, and 17.5%. Change detection between ALS-derived CHM (time 1) and DSI-derived CHM (time 2) provided change estimates that demonstrated good agreement (r _0.71) with two-date, ALS only, change outputs. For the single-layered, even-aged stands under investigation in this study, the DSI-derived vertical information is an appropriate and cost-effective data source for estimating and updating forest information. The accuracy of DSI information is based on a capability to measure the height of the upper canopy envelope with performance analogous to ALS. Forest attributes that are well captured and subsequently modeled from height metrics are best suited to estimation from DSI metrics, whereas ALS is more suitable for capturing stand density. Further investigation is required to better understand the performance of DSI-derived height products in more complex forest environments. Furthermore, the difference in variance captured between ALS and DSI-derived CHM also needs to be better understood in the context of change detection and inventory update considerations.
The forest canopy is one of the chief determinants of the microhabitat within the forest. It affects plant growth and survival, hence determining the nature of the vegetation, and wildlife habitat. A plethora of different techniques have been devised to measure the canopy. Evaluation of the literature reveals confusion over what is actually being measured. This paper distinguishes two basic types of measurement: canopy cover is the area of the ground covered by a vertical projection of the canopy, while canopy closure is the proportion of the sky hemisphere obscured by vegetation when viewed from a single point. The principal techniques used to measure canopy cover, canopy closure, and a number of related measures are described and discussed. The advantages and limitations are outlined and some sampling guidelines are provided. The authors hope to clarify the nature of the measurements and to provide foresters with sufficient information to select techniques suitable for their needs.
Airborne laser scanning is today considered the most accurate remote sensing method for forest inventory. The main advantage of laser scanning is the three-dimensional data. Three-dimensional canopy surface models can also be derived by means of digital aerial photogrammetry on the basis of optical remote sensing imagery. The photogrammetric surface models require high-resolution aerial images with stereo coverage. In this study, both a canopy height model derived from a photogrammetric digital surface model and laser point data were tested in estimation of sample-plot-level forest attributes. The attributes tested include diameter, mean and dominant height, basal area, and volume of growing stock. The results indicate that the laser data give higher accuracy for the estimated forest variables than does the photogrammetric canopy height model. The stand dominant height was the most accurately estimated variable from both data sources and showed the smallest difference between the laser data and photogrammetric canopy height models. The performance of the photogrammetric model was poorest in estimation of basal area and volume of growing stock.
Image matching is a key procedure in the process of generation of Digital Surface Models (DSM). We have developed a new approach for image matching and the related software package. This technique has proved its good performance in many applications. Here, we demonstrate its use in 3D tree modelling. After a brief description of our image matching technique, we show results from analogue and digital aerial images and high-resolution satellite images (IKONOS). In some cases, comparisons with manual measurements and/or airborne laser data have been performed. The evaluation of the results, qualitative and quantitative, indicate the very good performance of our matcher. Depending on the data acquisition parameters, the photogrammetric DSM can be denser than a DSM generated by laser, and its accuracy may be better than that from laser, as in these investigations. The tree canopy is well modelled, without smoothing of small details and avoiding the canopy penetration occurring with laser. Depending on the image scale, not only dense forest areas but also individual trees can be modelled.
Airborne Laser Scanning (ALS), also known as Light Detection and Ranging (LiDAR) enables an accurate three-dimensional characterization of vertical forest structure. ALS has proven to be an information-rich asset for forest managers, enabling the generation of highly detailed bare earth digital elevation models (DEMs) as well as estimation of a range of forest inventory attributes (including height, basal area, and volume). Recently, there has been increasing interest in the advanced processing of high spatial resolution digital airborne imagery to generate image-based point clouds, from which vertical information with similarities to ALS can be produced. Digital airborne imagery is typically less costly to acquire than ALS, is well understood by inventory practitioners, and in addition to enabling the derivation of height information, allows for visual interpretation of attributes that are currently problematic to estimate from ALS (such as species, health status, and maturity). At present, there are two limiting factors associated with the use of image-based point clouds. First, a DEM is required to normalize the image-based point cloud heights to aboveground heights; however DEMs with sufficient spatial resolution and vertical accuracy, particularly in forested areas, are usually only available from ALS data. The use of image-based point clouds may therefore be limited to those forest areas that already have an ALS-derived DEM. Second, image-based point clouds primarily characterize the outer envelope of the forest canopy, whereas ALS pulses penetrate the canopy and provide information on sub-canopy forest structure. The impact of these limiting factors on the estimation of forest inventory attributes has not been extensively researched and is not yet well understood. In this paper, we review the key similarities and differences between ALS data and image-based point clouds, summarize the results of current research related to the comparative use of these data for forest inventory attribute estimation, and highlight some outstanding research questions that should be addressed before any definitive recommendation can be made regarding the use of image-based point clouds for this application.
Ranging techniques such as lidar (LIght Detection And Ranging) and digital stereo‐photogrammetry show great promise for mapping forest canopy height. In this study, we combine these techniques to create hybrid photo‐lidar canopy height models (CHMs). First, photogrammetric digital surface models (DSMs) created using automated stereo‐matching were registered to corresponding lidar digital terrain models (DTMs). Photo‐lidar CHMs were then produced by subtracting the lidar DTM from the photogrammetric DSM. This approach opens up the possibility of retrospective mapping of forest structure using archived aerial photographs. The main objective of the study was to evaluate the accuracy of photo‐lidar CHMs by comparing them to reference lidar CHMs. The assessment revealed that stereo‐matching parameters and left–right image dissimilarities caused by sunlight and viewing geometry have a significant influence on the quality of the photo DSMs. Our study showed that photo‐lidar CHMs are well correlated to their lidar counterparts on a pixel‐wise basis (r up to 0.89 in the best stereo‐matching conditions), but have a lower resolution and accuracy. It also demonstrated that plot metrics extracted from the lidar and photo‐lidar CHMs, such as height at the 95th percentile of 20 m×20 m windows, are highly correlated (r up to 0.95 in general matching conditions).
The aim of the paper is to sum up knowledge of colour infrared (CIR) aerial photography as a tool for vegetation data for mapping and monitoring in environmental and biodiversity surveys and change detection surveillance. It compiles thirty years of research of the main ecosystems in Swedish vegetation, where the overall goal was to develop methods for mapping and monitoring vegetation by use of CIR aerial photographs, assess the accuracy compared to field-based mapping and to implement them as a tool in nature conservation and environmental planning. The methods include development of a classification system, identification and analysis of indicators, development of interpretation techniques, and evaluation compared to the data collected in the field. The CIR observable criteria are colour, texture, pattern, size, form, and density, based on spectral reflectance, physiognomy, life forms, ecological conditions, moisture and nutrition, vegetation period and phenology, topography, site conditions, and management methods. The methods have been used to produce vegetation maps of mountains, boreal forests, and mires in northern and central Sweden, in national inventories of wetlands, ancient meadows and pastures, key biotopes in forests and for monitoring agricultural landscapes.
This paper presents an approach to assess increase and decrease (2002–1997) of forest area and other wooded areas in a mire biotope in the Pre-alpine zone of Central Switzerland using logistic regression models and airborne remote sensing data (CIR aerial images, DSM derived from them). The present study was carried out in the framework of the Swiss Mire Protection Program, where increase and decrease of forest areas are key issues. In a first step, automatic DSMs were generated using an image matching approach from CIR aerial images of 1997 and 2002. In a second step, the DSMs were co-registered and normalized using LiDAR data. Tree layers from both years of various levels of detail were then generated combining canopy covers derived from normalized DSMs with a multi-resolution segmentation and a fuzzy classification. On the basis of these tree layers, fractional tree/shrub covers were calculated using explanatory variables derived from these DSMs only. Bias was estimated by analysing the distribution of the fractional model differences. The corrected models reveal a decrease of tree/shrub probability which indicates a decrease of forest and other wooded areas between 1997 and 2002. The models also indicate real shrub encroachment in open mire. The detection of shrub encroachment may be helpful for selective logging purposes for sustainable mire habitat management. The study stresses the importance of high-resolution and high-quality DSMs and highlights the potential of fractional covers for ecological modeling.
The landscape-level and multiscale biodiversity monitoring program National Inventory of Landscapes in Sweden (NILS) was launched in 2003. NILS is conducted as a sample-based stratified inventory that acquires data across several spatial scales, which is accomplished by combining aerial photo interpretation with field inventory. A total of 631 sample units are distributed across the land base of Sweden, of which 20% are surveyed each year. By 2007 NILS completed the first 5-year inventory phase. As the reinventory in the second 5-year phase (2008-2012) proceeds, experiences and insights accumulate and reflections are made on the setup and accomplishment of the monitoring scheme. In this article, the emphasis is placed on background, scope, objectives, design, and experiences of the NILS program. The main objective to collect data for and perform analyses of natural landscape changes, degree of anthropogenic impact, prerequisites for natural biological diversity and ecological processes at landscape scale. Different environmental conditions that can have direct or indirect effects on biological diversity are monitored. The program provides data for national and international policy and offers an infrastructure for other monitoring program and research projects. NILS has attracted significant national and international interest during its relatively short time of existence; the number of stakeholders and cooperation partners steadily increases. This is constructive and strengthens the incentive for the multiscale monitoring approach.
Remotely sensed high-resolution imagery and LiDAR data can be used to develop stand delineations and stratifications for forest inventory and management purposes. A new Area Dependent Region Merging method is introduced that uses LiDAR data and expert knowledge to develop forest stands and strata based on user-supplied constraints. This method uses an area-dependent scale parameter that allows for different merging criteria based on the size of the objects being merged. This method was used to develop a new forest inventory that showed improved accuracy with significantly fewer field plots. Compared to non-area-dependent region merging approaches, this method more effectively reduced within stand variability and more closely matched a manual stand delineation.
In this study, a novel automatic forest stand segmentation method based on Voronoi cells and Airborne Laser Scanner data was developed and validated using a systematic grid of field plots. The automatic method produce results comparable to manual stand delineation.
The rapid development in aerial digital cameras in combination with the increased availability of high-resolution Digital Elevation Models (DEMs) provides a renaissance for photogrammetry in forest management planning. Tree height, stem volume, and basal area were estimated for forest stands using canopy height, density, and texture metrics derived from photogrammetric matching of digital aerial images and a high-resolution DEM. The study was conducted at a coniferous hemi-boreal site in southern Sweden. Three different data-sets of digital aerial images were used to test the effects of flight altitude and stereo overlap on an area-based estimation of forest variables. Metrics were calculated for 344 field plots (10 m radius) from point cloud data and used in regression analysis. Stand level accuracy was evaluated using leave-one-out cross validation of 24 stands. For these stands the tree height ranged from 4.8 to 26.9 m (17.8 m mean), stem volume 13.3 to 455 m3 ha−1 (250 m3 ha−1 mean), and basal area from 4.1 to 42.9 m2 ha−1 (27.1 m2 ha−1 mean) with mean stand size of 2.8 ha. The results showed small differences in estimation accuracy of forest variables between the data-sets. The data-set of digital aerial images corresponding to the standard acquisition of the Swedish National Land Survey (Lantmäteriet), showed Root Mean Square Errors (in percent of the surveyed stand mean) of 8.8% for tree height, 13.1% for stem volume and 14.9% for basal area. The results imply that photogrammetric matching of digital aerial images has significant potential for operational use in forestry.
In a collaboration between the Swedish Space Corporation and the National Land Survey of Sweden, work has been carried out to develop a satellite-data-based method for mapping wetlands for the Swedish CORINE Land Cover database. The aim was two-fold: (1) develop a method for mapping the wetlands in five classes, namely "inland marshes", "wet mires", "other mires", "exploited peatlands", and "salt marshes", with a minimum mapping unit of 1 ha; and (2) investigate if the classes "wet mires" and "other mires" could be obtained more reliably from satellite data than directly from 1: 50 000 scale topographic maps. Data sources used in the project included geo-referenced Landsat thematic mapper (TM) images, digital topographic map data, digital vegetation data, and colour infrared aerial photographs. The development work included analysis of spectral signature characteristics, classification tests, method definition, method verification, pilot production, and accuracy assessment. The method developed to obtain the five wetland classes is a combination method. The classes inland marshes, salt marshes, and exploited peatlands are mapped using geographical information system (GIS) operations, combining information from the topographic maps, and visual interpretation of Landsat TM data. For the classes wet mires and other mires, a stepwise semi-automatic approach was adopted, where level slicing in ratios or in single Landsat TM bands is used to produce the classes. The evaluation showed high accuracy for all five classes, with overall accuracies ranging from 90% to 100%. It was also concluded that the semi-automatic interactive approach used to produce wet mires and other mires gave the same high accuracy as the topographic map, or a better result than the map for areas in northern Sweden where the representation of wet mires on the map is not as good, The method described is used for production of the wetland classes for the Swedish CORINE Land Cover.
Recent research results have shown that the performance of digital
surface model extraction using novel high-quality photogrammetric images
and image matching is a highly competitive alternative to laser
scanning. In this article, we proceed to compare the performance of
these two methods in the estimation of plot-level forest variables.
Dense point clouds extracted from aerial frame images were used to
estimate the plot-level forest variables needed in a forest inventory
covering 89 plots. We analyzed images with 60% and 80% forward overlaps
and used test plots with off-nadir angles of between 0° and 20°.
When compared to reference ground measurements, the airborne laser
scanning (ALS) data proved to be the most accurate: it yielded root mean
square error (RMSE) values of 6.55% for mean height, 11.42% for mean
diameter, and 20.72% for volume. When we applied a forward overlap of
80%, the corresponding results from aerial images were 6.77% for mean
height, 12.00% for mean diameter, and 22.62% for volume. A forward
overlap of 60% resulted in slightly deteriorated RMSE values of 7.55%
for mean height, 12.20% for mean diameter, and 22.77% for volume.
According to our results, the use of higher forward overlap produced
only slightly better results in the estimation of these forest
variables. Additionally, we found that the estimation accuracy was not
significantly impacted by the increase in the off-nadir angle. Our
results confirmed that digital aerial photographs were about as accurate
as ALS in forest resources estimation as long as a terrain model was
available.
The objectives of this study were to identify useful predictive factors for tree species identification of individual trees and to compare classifications based on a combination of LiDAR data and multi‐spectral images with classification by the use of each individual data source. Crown segments derived from LiDAR data were mapped to multi‐spectral images for extraction of spectral data within individual tree crowns. Several features, related to height distribution of laser returns in the canopy, canopy shape, proportion of different types of laser returns, and intensity of laser returns, were derived from LiDAR data. Data from a test site in southern Sweden were used (lat. 58°30′ N, long. 13°40′ E). The forest consisted of Norway spruce (Picea abies), Scots pine (Pinus sylvestris), and deciduous trees. Classification into these three tree species groups was validated for 1711 trees that had been detected in LiDAR data within 14 field plots (sizes of 20×50 m or 80×80 m). The LiDAR data were acquired by the TopEye MkII system (50 LiDAR measurements per m) and the multi‐spectral images were taken by the Zeiss/Intergraph Digital Mapping Camera. The overall classification accuracy was 96% when both data sources were combined.
A modification to the maximum likelihood algorithm was developed for classification of forest types in Sweden's part of the CORINE land cover mapping project. The new method, called the “calibrated maximum likelihood classification” involves an automated and iterative adjustment of prior weights until class frequency in the output corresponds to class frequency as calculated from objective (field-inventoried) estimates. This modification compensates for the maximum likelihood algorithm's tendency to over-represent dominant classes and under-represent less frequent ones. National forest inventory plot data measured from a five-year period are used to estimate relative frequency of class occurrence and to derive spectral signatures for each forest class. The classification method was implemented operationally within an automated production system which allowed rapid production of a country-wide forest type map from Landsat TM/ETM+ satellite data. The production system automated the retrieval and updating of forest inventory plots, a plot-to-image matching routine, illumination and haze correction of satellite imagery, and classification into forest classes using the calibrated maximum likelihood classification. This paper describes the details of the method and demonstrates the result of using an iterative adjustment of prior weights versus unadjusted prior weights. It shows that the calibrated maximum likelihood algorithm adjusts for the overclassification of classes that are well represented in the training data as well as for other classes, resulting in an output where class proportions are close to those as expected based on forest inventory data.
This paper deals with the production of digital surface models (DSM) from high resolution images. The paper explains why the quality of the dense DSM depends on the quality of input data and data compilation. The INPHO GmbH software MATCH-T DSM has been redesigned to produce very dense DSM data. The most important improvement was the introduction of sequential multi-image matching. The point extraction is no longer based on static models, but on computation units. Each computation unit in MATCH-T DSM chooses the best suited image pairs. Each image pair delivers a point cloud, the combined point clouds are filtered by a robust analysis. The precision and the completeness of the MATCH-T DSM from high resolution images is analyzed in two case studies. RESUME: Cet article a pour sujet la production de modèles numériques de surface (MNS) à partir d'images aériennes à haute résolution. L'article montre pourquoi la qualité du MNS dense dépend fortement de la qualité des données d'entrée et du mode d'acquisition. Le logiciel MATCH-T DSM développé par INPHO GmbH a été reconçu pour pouvoir produire des MNS très denses. Pour ce faire, une méthode d'autocorrélation séquentielle a été développée. L'extraction n'est plus effectuée par une compilation de modèles stéréoscopiques statiques ; pour chaque unité d'extraction, MATCH-T DSM choisit les paires d'images les plus appropriées ; chaque paire d'images fournit un nuage de points qui sont ensuite filtrés à l'aide d'une analyse statistique robuste. La précision et la complétude du MNS extrait avec MATCH-T DSM sont évaluées dans deux études.
Question: What precision and accuracy of visual cover estimations can be achieved after repeated calibration with images of vegetation in which the true cover is known, and what factors influence the results?
Methods: Digital images were created, in which the true cover of vegetation was digitally calculated. Fifteen observers made repeated estimates with immediate feedback on the true cover. The effects on precision and accuracy through time were evaluated with repeated proficiency tests. In a field trial, cover estimates, before and after calibration, were compared with point frequency data.
Results: Even a short time of calibration greatly improves precision and accuracy of the estimates, and can also reduce the influence of different backgrounds, aggregation patterns and experience. Experienced observers had a stronger tendency to underestimate the cover of narrow-leaved grasses before calibration. The field trial showed positive effects of computer-based calibration on precision, in that it led to considerably less between-observer variation for one of the two species groups.
Conclusions: Computer-aided calibration of vegetation cover estimation is simple, self-explanatory and time-efficient, and might possibly reduce biases and drifts in estimate levels over time. Such calibration can also reduce between-observer variation in field estimates, at least for some species. However, the effects of calibration on estimations in the field must be further evaluated, especially for multilayered vegetation.
Photogrammetric methods using parallaxes can be employed to measure tree heights on aerial photographs. Because it is often impossible to measure ground elevation near trees growing in dense forests, such height measurements remain prone to error. Our objective was to solve this problem by combining a stereomodel and a digital terrain model (DTM) produced by an airborne-scanning system that uses light detection and ranging (lidar). A stereopair of scanned aerial photographs was first registered to a lidar DTM. The elevation of the apex of 202 Thuja occidentalis (L.) individuals was measured by an observer on a digital photogrammetric workstation. The tree base elevations were read from the lidar DTM and subtracted from the corresponding apex elevations to calculate individual tree heights. These were then compared with the heights measured in the field. The average photo-lidar bias was 0.59 m, and the average deviation of 1.01 m decreased to 0.88 m when the bias was removed. It was demonstrated that the photographic clearness of the tree apices influences the height error, while the density of the lidar echoes under the forest canopy does not. Using this method, retrospective studies of changes in tree height become feasible by using archived aerial photographs and recent lidar DTMs.
Various studies have been presented within the last 10 years on the possibilities for predicting forest variables such as stand volume and mean height by means of airborne laser scanning (ALS) data. These have usually considered tree stock as a whole, even though it is tree species-specific forest information that is of primary interest in Finland, for example. We will therefore concentrate here on prediction of the species-specific forest variables volume, stem number, basal area, basal area median diameter and tree height, applying the non-parametric k-MSN method to a combination of ALS data and aerial photographs in order to predict these stand attributes simultaneously for Scots pine, Norway spruce and deciduous trees as well as total characteristics as sums of the species-specific estimates. The predictor variables derived from the ALS data were based on the height distribution of vegetation hits, whereas spectral values and texture features were employed in the case of the aerial photographs. The data covered 463 sample plots in 67 stands in eastern Finland, and the results showed that this approach can be used to predict species-specific forest variables at least as accurately as from the current stand-level field inventory for Finland. The characteristics of Scots pine and Norway spruce were predicted more accurately than those of deciduous trees.
This paper describes the use of aerial photography and airborne LiDAR to estimate individual tree heights in forest stands. The advantages and disadvantages in the use of LiDAR systems are revised and a data fusion analysis with digital aerial photography is proposed. An example of the use of these techniques in a forested area in Scotland is presented. An algorithm has been developed to extract a high-resolution digital terrain model of the bare ground. This provided a tree canopy model as the difference between the laser first pulse and the model of the underlying terrain. Information about individual trees was obtained by image segmentation and classification. This analysis provided a good method of estimating individual tree canopies and heights.
A comparison between data acquisition and processing from passive optical sensors and airborne laser scanning is presented. A short overview and the major differences between the two technologies are outlined. Advantages and disadvantages with respect to various aspects are discussed, like sensors, platforms, flight planning, data acquisition conditions, imaging, object reflectance, automation, accuracy, flexibility and maturity, production time and costs. A more detailed comparison is presented with respect to DTM and DSM generation. Strengths of laser scanning with respect to certain applications are outlined. Although airborne laser scanning competes to a certain extent with photogrammetry and will replace it in certain cases, the two technologies are fairly complementary and their integration can lead to more accurate and complete products, and open up new areas of application.
Airborne laser scanning systems are opening new possibilities for surveys and documentation of difficult areas and objects, such as dense city areas, forest areas and electrical power lines. Laser scanner systems available on the market are presently in a fairly mature state of art while the processing of airborne laser scanner data still is in an early phase of development. To come from irregular 3D point clouds to useful representations and formats for an end-user requires continued research and development of methods and algorithms for interpretation and modelling. This paper presents some methods and algorithms concerning filtering for determining the ground surface, DEM, classification of buildings for 3D City Models and the detection of electrical power lines. The classification algorithms are based on the Minimum Description Length criterion. The use of reflectance data and multiple echoes from the laser scanner is examined and found to be useful in many applications.
Remote sensing from airborne and spaceborne platforms provides valuable data for mapping, environmental monitoring, disaster management and civil and military intelligence. However, to explore the full value of these data, the appropriate information has to be extracted and presented in standard format to import it into geo-information systems and thus allow efficient decision processes. The object-oriented approach can contribute to powerful automatic and semi-automatic analysis for most remote sensing applications. Synergetic use to pixel-based or statistical signal processing methods explores the rich information contents. Here, we explain principal strategies of object-oriented analysis, discuss how the combination with fuzzy methods allows implementing expert knowledge and describe a representative example for the proposed workflow from remote sensing imagery to GIS. The strategies are demonstrated using the first object-oriented image analysis software on the market, eCognition, which provides an appropriate link between remote sensing imagery and GIS.
In this paper, we present a two-stage approach for characterizing the structure of Pinus sylvestris L. stands in forests of central Spain. The first stage was to delimit forest stands using eCognition and a digital canopy height model (DCHM) derived from lidar data. The polygons were then clustered (k-means algorithm) into forest structure types based on the DCHM data within forest stands. Hypsographs of each polygon and field data validated the separability of structure types. In the study area, 112 polygons of Pinus sylvestris were segmented and classified into five forest structure types, ranging from high dense forest canopy (850 trees ha−1 and Loreýs height of 17.4 m) to scarce tree coverage (60 tree ha−1 and Loreýs height of 9.7 m). Our results indicate that the best variables for the definition and characterization of forest structure in these forests are the median and standard deviation (S.D.), both derived from lidar data. In these forest types, lidar median height and standard deviation (S.D.) varied from 15.8 m (S.D. of 5.6 m) to 2.6 m (S.D. of 4.5 m). The present approach could have an operational application in the inventory procedure and forest management plans.
This paper presents analyses of different methods of postprocessing lines that have resulted from the raster-to-vector conversion
of black and white line drawing. Special attention was paid to the borders of connected components of maps. These methods
are implemented with compression and smoothing algorithms. Smoothing algorithms can enhance accuracy, so using both smoothing
and compression algorithms in succession gives a more accurate result than using only a compression algorithm. The paper also
shows that a map in vector format may require more memory than a map in raster format. The Appendix contains a detailed description
of the new smoothing method (continuous local weighted averaging) suggested by the authors.
Derivation of Forest Parameters from UltracamD Data.” PhD diss
- M Hirschmugl
Manual for Aerial Photo Interpretation in the National Inventory of Landscapes in Sweden
- A Allard
- B Nilsson
- K Pramborg
- G Ståhl
- S Sundquist
Match-T DSM 5.2 Reference Manual
- Anonymous
Geographical Vegetation Data of Lantmäteriet in Sweden Ås: Norsk institutt for skog og landskap
- L Andersson
Produktbeskrivning: Digitala Flygbilder. Version 1.2. GSD Geografiska Sverigedata™ Gävle: Lantmäteriet. http://www.lantmateriet.se/Global
- Anonymous
DEM Generation from Laser Scanner Data Using Adaptive TIN Models.” International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XXXIII
- P Axelsson
Forest Statistics 2013. Official Statistics of Sweden
- P Nilsson
- N Cory
- J Fridman
- G Kempe
TerraScan for MicroStation, User’s Guide
- A Soininen