Conference PaperPDF Available

DTM Generation in Forest Regions From Satellite Stereo Imagery

Authors:
Jiaojiao Tian at German Aerospace Center (DLR)
Jiaojiao Tian
Thomas Krauß at German Aerospace Center (DLR)
Thomas Krauß
Peter Reinartz at German Aerospace Center (DLR)
Peter Reinartz
Download full-text PDF
Read full-text
Download citation

Abstract and Figures

Satellite stereo imagery is becoming a popular data source for derivation of height information. Many new Digital Surface Model (DSM) generation and evaluation methods have been proposed based on these data. A novel Digital Terrain Model (DTM) extraction method based on the DSM from satellite stereo imagery is proposed in this paper. Instead of directly filtering the DSM, firstly a single channel based classification method is proposed. In this step, no multi-spectral information is used, because for some stereo sensors, like Cartosat-1, only panchromatic channels are available. The proposed classification method adopts the random forests method to get initial probability maps of the four main classes in forest regions (high-forest, low-forest, ground, and buildings). To cover the pepper and salt effect of this pixel based classification method, the probability maps are further filtered based on the adaptive Wiener filtering. Then a cube-based greedy strategy is applied in generating the final classification map from these refined probability maps. Secondly, the height distances between neighboring regions are calculated along the boundary regions. These height distances can be used to estimate the relative region heights. Thirdly, the DTM is extracted by subtracting these relative region heights from the DSM in the order of: buildings – low forest – high forest. In the end, the extracted DTM is further smoothed using median filter. The proposed DTM extraction method is finally tested on satellite stereo imagery captured by Cartosat-1. Quality evaluation is performed by comparing the extracted DTMs to a reference DTM, which is generated from the last return airborne laser scanning point cloud.
Figures - uploaded by Peter Reinartz
Author content
All figure content in this area was uploaded by Peter Reinartz
Content may be subject to copyright.
Content uploaded by Peter Reinartz
Author content
All content in this area was uploaded by Peter Reinartz on Dec 09, 2014
Content may be subject to copyright.
DTM GENERATION IN FOREST REGIONS FROM SATELLITE STEREO IMAGERY
J. Tian, T. Krauss, P. Reinartz
Remote Sensing Technology Institute, German Aerospace Center (DLR), 82234 Wessling, Germany
(Jiaojiao.Tian, Thomas.Krauss, Peter.Reinartz)@dlr.de
Commission I, WG I/4
KEY WORDS: Optical Stereo Data, DTM, DSM, Forest, Random Forests
ABSTRACT:
Satellite stereo imagery is becoming a popular data source for derivation of height information. Many new Digital Surface Model
(DSM) generation and evaluation methods have been proposed based on these data. A novel Digital Terrain Model (DTM)
extraction method based on the DSM from satellite stereo imagery is proposed in this paper. Instead of directly filtering the
DSM, firstly a single channel based classification method is proposed. In this step, no multi-spectral information is used, because
for some stereo sensors, like Cartosat-1, only panchromatic channels are available. The proposed classification method adopts the
random forests method to get initial probability maps of the four main classes in forest regions (high-forest, low-forest, ground,
and buildings). To cover the pepper and salt effect of this pixel based classification method, the probability maps are further
filtered based on the adaptive Wiener filtering. Then a cube-based greedy strategy is applied in generating the final classification
map from these refined probability maps. Secondly, the height distances between neighboring regions are calculated along the
boundary regions. These height distances can be used to estimate the relative region heights. Thirdly, the DTM is extracted by
subtracting these relative region heights from the DSM in the order of: buildings - low forest - high forest. In the end, the
extracted DTM is further smoothed using median filter.
The proposed DTM extraction method is finally tested on satellite stereo imagery captured by Cartosat-1. Quality evaluation is
performed by comparing the extracted DTMs to a reference DTM, which is generated from the last return airborne laser scanning
point cloud.
1. INTRODUCTION
Digital Terrain Models (DTMs) provide essential
information for many remote sensing projects and
applications. A DTM can be extracted from a Digital
Surface Model (DSM) automatically by removing non-
ground objects and filtering these areas with proper
elevation values (Kraus and Pfeifer, 2001; Arefi et al.,
2011; Krauss et al., 2011). When both models DSM and
DTM are available, the absolute height of buildings and
trees can be calculated.
Various DTM generation methods have been proposed
when different sources of DSMs are used. Photogrammetry
and laser scanning are still two of the most important
approaches in DSM/DTM generation. A detailed
description and comparison between photogrammetry and
airborne laser scanning (ALS) approaches can be found in
Baltsavias (1999). Generally, ALS data are more advanced
for DTM generation. In urban regions, ALS data are more
accurate and building boundaries are modeled better. In
forest regions, most of the ALS last return points show
already the ground pixels, and trees are well separated from
each other. A DTM can be generated by filtering these
remaining trees.
Compared to ALS data, the DSMs generated from satellite
stereo imagery (stereo-DSM) are more economical data
sources and also advantageous for larger region survey and
mapping applications (Zhang et al., 2005; d’Angelo et al.,
2010; Tian et al., 2013). Moreover, the resolution as well as
the quality of the extracted DSMs is improving (Straub et
al., 2013). However, not many specific DTM extracting
methods have been proposed for these data. Directly
filtering the DSMs (Pfeifer et al., 2001; Arefi et al., 2011)
will not work so well in dense forest regions, since in the
stereo-DSMs, ground height between trees cannot be
observed.
In this paper, a novel DTM generation method is proposed
specifically for stereo-DSM in forest regions. Instead of
directly filtering the DSM, the original satellite image is
used to provide an initial classification result. To cover the
pepper and salt effect of this pixel based classification
method, a refined random forest classification method is
introduced. Besides classification, in the second section, the
step down DTM interpretation workflow is described. The
experiments and results are shown and evaluated in the
third section. The last section is conclusion.
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-1, 2014
ISPRS Technical Commission I Symposium, 17 – 20 November 2014, Denver, Colorado, USA
This contribution has been peer-reviewed.
doi:10.5194/isprsarchives-XL-1-401-2014
401
2. METHODS
A three steps DTM generation procedure is proposed in this
paper. As shown in Fig. 1, in the first step, the orthorectifed
satellite panchromatic images are used to get an initial
classification result. Based on the classification results, the
terrain height can be calculated from lower objects to
higher objects. In the end, a further refinement of the DTM
is applied.
Panchromatic
image
High forest
DSM
Low forest
Ground
Buildings
Classification
DTM1
DTM2
DTM3
Refinement
DTM
Figure 1. Flow chart of the proposed method
2.1 Classification
Random Forest is a robust and powerful machine learning
classifier, which is capable of processing large datasets
(Breiman, 2001). The random forest consists of several
decision trees, also called classification trees. These
classification trees are “grown” based on training sets.
Random forest builds the tree nodes randomly with random
features. Each decision tree is used separately to classify
the satellite image of interest. By summarizing these
classification results, one classification map and several
membership maps for each class can be obtained.
In this step, no multi-spectral information is used, because
for some stereo sensors, like Cartosat-1, only panchromatic
channels are available. The proposed classification method
adopts the multi-level features explained in Tian et. al
(2013) to get initial probability maps of the four main
classes in forest regions. These four classes are high-forest,
low-forest, ground, and buildings.
To overcome the drawback of the pixel based classification
method, which leads to a salt-and-pepper effect, two
approaches are considered in this step. Besides the
classification result, random forest can also provide the
probability map of all classes. Therefore, an adaptive
Wiener filtering is applied to improve the quality of each
probability map. Afterwards, instead of only considering
the probability sets pixel wise, their neighborhood pixels
are also considered in the decision making procedure. It can
be called ‘cube-based greedy strategy’. The class label
which has the highest probability in this probability cube is
selected as the label of the pixel.
2.2 Step down DTM generation
Different to urban regions, in forest regions large areas of
higher objects are connected together, which makes the
DTM generation in forest regions more difficult. In this
paper, we tried to separate large forest regions into a
number of small regions by using the classification result.
The relative heights of three high-level object classes are
calculated respectively.
As shown in Figure 2, in each step, the height distance
between the higher objects and lower objects can be
calculated. The lower forest and higher forest canopy are
shown in green and light blue color respectively. The red
point represents one obtained boundary pixels. The yellow
rectangle is the pre-defined search window. We take the
height distance inside this search window as the step height
of that red pixel. For each high object region, the object
height is obtained by calculating the average step height
ΔH of all boundary pixel, which means the height distance
of h1 and h2.
The DTM is extracted by subtracting these relative region
heights from the DSM in the order of: buildings - low
forest - high forest. The building class is processed as the
first step, as it is mainly separated from the forest region.
Figure 2 focuses on explaining the DTM generation for low
/ high forest areas.
(a) (Step1)
(b)
(c)
Figure 2. Step down height calculation method (green: low
forest, blue: high forest; black: ground surface)
h
1
h
2
h
2
´
h
1
´
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-1, 2014
ISPRS Technical Commission I Symposium, 17 – 20 November 2014, Denver, Colorado, USA
This contribution has been peer-reviewed.
doi:10.5194/isprsarchives-XL-1-401-2014
402
2.3 DTM refinement
A further refinement of the generated DTM is necessary.
Since the regions in each class have been processed
separately, there would be a height difference remaining in
the boundary region. Therefore a median filtering is
adopted to improve the quality of the generated DSM.
3. EXPERIMENT AND RESULTS
3.1 Datasets
In this paper, Cartosat-1 stereo imagery captured in
Traunstein, Germany are selected for the experiment.
Figure 3a is the panchromatic image. The generated DSMs
from Cartosat-1 are displayed in Figure 3b and 3c
respectively. The DSM generation procedure and quality
evaluation have been described in Tian et al. (2013) and
Straub et al. (2013).
To validate the accuracy and efficiency of the proposed
method the DTM from Laser scanning data is used as
reference data (shown in Figure 3c). All Cartosat-1 images
have been resampled to 2.5 meter resolution.
(a)
(b)
(d)
Figure 3. Experimental datasets: (a) panchromatic image;
(b) DSM from Cartosat-1; (c) reference DTM
3.2 Results
3.2.1 Classification
The classification result for Carosat-1 is shown in Figure
4b. The tour classes are marked with red, light green, dark
green and grey respectively. In order to show the
improvement, the original classification result from random
forests is displayed in Figure 4a. By referring to Figure 3a,
the original random forest can label these four classes
correctly. Limited to the salt-and-pepper effect, DTM
generation cannot be easily performed. The proposed
classification result (shown in Figure 4b) is visually much
better. Different classes are well separately from each
other. It can be observed that high forest regions are mainly
surrounded by lower forest. Though some roads are
wrongly classified as buildings, we will process them in the
first step. Since they do not have height distance to the
ground, they will keep their original height in DTM1.
(a)
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-1, 2014
ISPRS Technical Commission I Symposium, 17 – 20 November 2014, Denver, Colorado, USA
This contribution has been peer-reviewed.
doi:10.5194/isprsarchives-XL-1-401-2014
403
(b)
Figure 4. Comparison of the original classification map (a)
and the refined classification results (b). (Red: buildings;
Grey: ground; Dark green: high forest; Light green: low
forest).
3.2.2 DTM generation
In the second part we evaluate the behavior of the DTM
generation method. Two regions in the test data are
analyzed, one is a forest region with high/low forest, and
the other has a house in the middle of it. The results from
Cartosat-1 in each step are recorded in Figure 5 and Figure
6 respectively. For the forest region, we have displayed
here all three DTMs in the generation procedure and the
final refined DTM (Figure 5f). Figure 5g is the reference
DTM. In Figure 6, since no forest exists around that house,
only DTM1 is generated.
(a) (b) (c)
(d) (e) (f)
(g)
Figure 5. DTM generation procedure of the forest example
(a) panchromatic images ; (b) original DSM; (c) DTM1; (d)
DTM2; (e)DTM3; (f) Refined DTM; (g) reference DTM.
(a) (b) (c)
(d) (e)
Figure 6. DTM generation procedure of building example
(a) panchromatic images ; (b) original DSM; (c) DTM1; (d)
refined DTM; (e) reference DTM.
To prove the advantage of our proposed method, we have
compared our generated DTMs to the ones generated with
tophat morphological filtering. They are both compared to
the reference DTM.
Figure 7 and 8 show the profile comparison of the DSM
and DTMs of the forest and buildings respectively. For
better description, the profiles of the original DSMs are
also shown in these two figures with black color. DTMs
generated from our method and tophat have been displayed
with red and blue color respectively. The reference DTM is
shown with green color. As can be seen, for normal
building / house, the extracted DTM matches well with the
reference DTM. In the forest region, though the obtained
DTM has some height difference from the reference DTM,
the terrain slopes are well preserved.
Figure 7. Profile comparison between the DTMs along the
red line in Figure 5a (Black: original DSM; Red: generated
DSM; Blue: tophat morphological reconstruction; Green:
reference DTM).
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-1, 2014
ISPRS Technical Commission I Symposium, 17 – 20 November 2014, Denver, Colorado, USA
This contribution has been peer-reviewed.
doi:10.5194/isprsarchives-XL-1-401-2014
404
Figure 8. Profile comparison between the DTMs along the
red line in Figure 6a (Black: original DSM; Red: generated
DSM; Blue: tophat morphological reconstruction; Green:
reference DTM).
4. CONCLUSION
In this paper, a new DTM generation approach is presented.
In contrast to the classical approaches, the proposed
method has fully used all information provided by the
stereo satellite imagery. As the first contribution, an easy
implementable classification refinement method is
proposed. Since not enough spectral features are available
for some stereo satellite images, we have used the multi-le
el texture information for the classification.
Based on the classification result, a region based step-down
DTM generation procedure is proposed. In our approach,
we suppose that trees in the same region exhibit the same
or similar height. The generated DSM can preserve the
original terrain information. However, in some cases, the
trees in the middle of the region can be higher than other
trees. Therefore, the generated DTM is much higher than
the reference DTM. This problem will be solved in the
future work.
Reference:
Arefi, H., d’Angelo, P., Mayer, H., and Reinartz, P., 2011.
Iterative approach for efficient digital terrain model
production from CARTOSAT-1 stereo images. Journal of
Applied Remote Sensing, vol. 5, no.1, pp. 1-19.
Baltsavias, E. P., 1999. A comparison between
photogrammetry and laser scanning, ISPRS Journal of
photogrammetry and Remote Sensing, vol. 54, no. 2, pp.
83-94.
Breiman, L., 2001. Random Forest. Machine Learning, vol.
45, no. 1, pp. 5-32.
d’Angelo, P., 2010. Image matching and outlier removal
for large scale DSM generation. Convergence in
Geomatics. CGC & ISPRS. ISPRS Symposium
Commission I, 15.-18. June 2010, Calgary, Canada.
Kraus, K. and Pfeifer, N., 2001. Advanced DTM generation
from LIDAR data. International Archives of
Photogrammetry Remote Sensing and Spatial Information
Sciences, vol. 34, no. 3/W4, pp. 23-30.
Krauss, T., Arefi, H. and Reinartz, P., 2011. Evaluation of
selected methods for extracting digital terrain models from
satellite born digital surface models in urban areas.
International Conference on Sensors and Models in
Photogrammetry and Remote Sensing (SMPR 2011), pp.1 -
7.
Pfeifer, N., Stadler, P., Briese, C., 2001. Derivation of
digital terrain models in the SCOP++ environment.
Proceedings of OEEPE Workshop on Airborne
Laserscanning and Interferometric SAR for Detailed
Digital Elevation Models, Stockholm, Sweden (2001)
(13p.)
Razak, K. A., Santangelo, M., Westen, C. J. V., Straatsma,
M. W. and de Jong, S. M., 2013. Generating an optimal
DTM from airborne laser scanning data for landslide
mapping in a tropical forest environment. Geomorphology,
vol. 190, pp. 112-125.
Straub, C., Tian, J., Seitz, R., and Reinartz, P., 2013.
Assessment of Cartosat-1 and WorldView-2 stereo imagery
in combination with a LiDAR DTM for timber volume
estimation in a highly structured forest in Germany.
Forestry, vol. 86, pp. 463-473.
Tian, J., Reinartz, P., d’Angelo, P. and Ehlers, M., 2013.
Region-based automatic building and forest change
detection on Cartosat-1 stereo imagery. ISPRS Journal of
Photogrammetry and Remote Sensing, vol. 79, pp. 229-
239.
Zhang, L., 2005. Automatic Digital Surface Model (DSM)
Generation from Linear Array Images. PhD Thesis, Swiss
Federal Institute of Technology Zürich, Switzerland.
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-1, 2014
ISPRS Technical Commission I Symposium, 17 – 20 November 2014, Denver, Colorado, USA
This contribution has been peer-reviewed.
doi:10.5194/isprsarchives-XL-1-401-2014
405
... [29], in separating non-ground objects from lidar-based DSMs [27]. Tian et al. (2014) used RF to classify the study region into different ground and non-ground classes based on multi-level texture features extracted from Cartosat-1 stereo images [30]. Noticeably, though indicating good capabilities in separating ground and non-ground observations, these ground filtering methods were mainly applied in urban scenarios. ...
... [29], in separating non-ground objects from lidar-based DSMs [27]. Tian et al. (2014) used RF to classify the study region into different ground and non-ground classes based on multi-level texture features extracted from Cartosat-1 stereo images [30]. Noticeably, though indicating good capabilities in separating ground and non-ground observations, these ground filtering methods were mainly applied in urban scenarios. ...
... Our results demonstrate the superior performance of the GMM-based method over K-means and filtering-based ground identification methods by locating the highest-quality ground samples. While more advanced supervised classification algorithms (e.g., support vector machine [61], random forest [30], and deep convolutional neural networks [62]) can be used to classify ground and non-ground from optical imagery, they are not suitable for forested sites in high-latitude regions where training samples are rarely available. Our GMM-based method is unsupervised and has great potential for high latitude regions covered by ArcticDEM. ...
Reconstructing Digital Terrain Models from ArcticDEM and WorldView-2 Imagery in Livengood, Alaska
Article
Full-text available
  • Apr 2023
ArcticDEM provides the public with an unprecedented opportunity to access very high-spatial resolution digital elevation models (DEMs) covering the pan-Arctic surfaces. As it is generated from stereo-pairs of optical satellite imagery, ArcticDEM represents a mixture of a digital surface model (DSM) over a non-ground areas and digital terrain model (DTM) at bare grounds. Reconstructing DTM from ArcticDEM is thus needed in studies requiring bare ground elevation, such as modeling hydrological processes, tracking surface change dynamics, and estimating vegetation canopy height and associated forest attributes. Here we proposed an automated approach for estimating DTM from ArcticDEM in two steps: (1) identifying ground pixels from WorldView-2 imagery using a Gaussian mixture model (GMM) with local refinement by morphological operation, and (2) generating a continuous DTM surface using ArcticDEMs at ground locations and spatial interpolation methods (ordinary kriging (OK) and natural neighbor (NN)). We evaluated our method at three forested study sites characterized by different canopy cover and topographic conditions in Livengood, Alaska, where airborne lidar data is available for validation. Our results demonstrate that (1) the proposed ground identification method can effectively identify ground pixels with much lower root mean square errors (RMSEs) (<0.35 m) to the reference data than the comparative state-of-the-art approaches; (2) NN performs more robustly in DTM interpolation than OK; (3) the DTMs generated from NN interpolation with GMM-based ground masks decrease the RMSEs of ArcticDEM to 0.648 m, 1.677 m, and 0.521 m for Site-1, Site-2, and Site-3, respectively. This study provides a viable means of deriving high-resolution DTM from ArcticDEM that will be of great value to studies focusing on the Arctic ecosystems, forest change dynamics, and earth surface processes.
View
... Compared to several others being tested, the inverse distance weighting (IDW) showed the best results. As a final step, image differencing of the DSM data was conducted resulting in a layer showing the changes in the terrain surface during the time of investigation (Tian et al. 2014a, Tian 2013, Zhang 2005. Although Tian et al. (2014a, b) demonstrated optimized results applying a robust differencing method, in this eastern DRC area, a pixel-wise differencing provided satisfying results matching the objectives of the underlying study. ...
... 2.5D change detection was carried out based on the DSMs of the four dates (see Table 1 stereo data). Changes in the terrain surface were generated by image differencing of the DSM data (Zhang, 2005;Tian, 2013;Tian et al., 2013;Tian et al., 2014a) (see Fig. 2h). The quality of the detected surface changes depend on the quality of input data, climatic conditions during data acquisition (e.g. ...
Dissertation Olaf Kranz AUG2018
Thesis
Full-text available
  • Aug 2018
The number of violent conflicts worldwide has tripled since 2010 and sub-Saharan Africa covering the Central African Republic, eastern Democratic Republic of the Congo, Burundi, South Sudan and Darfur (United Nations & Worldbank 2017, The Week 2016) is still a hotspot. The root causes of these violent regional conflicts are complex with partly reinforcing cycles, and a major result is forced migration affecting more than 65 million people by the end of 2015 (UNHCR 2016). Amongst the regions with highest amounts of conflict-related IDPs during the past decade are Sudan, South Sudan and the Democratic Republic of the Congo (DRC). Most migrants gather in camps or informal settlements causing increasing pressure on natural resources which in turn further intensifies local conflicts (HIIK 2013). In 2009 the United Nations Environment Programme (UNEP) Expert Advisory Group on Environment, Conflict and Peacebuilding stated that “there is significant potential for conflict over natural resources to intensify in the coming decades” (UNEP 2009). On the other hand, resource abundance might also lead to violent conflict by financing weapons through revenues from resource exploitation, a phenomenon which has become known as “resource curse” (Le Billon 2001, Mildner et al. 2011). In late 2008, the UN and the EU agreed on jointly building capacities to prevent land and natural resources from fuelling conflict (UN-HABITAT 2012, UNDP 2011). In order to mitigate the consequences of resulting humanitarian crises a large number of governmental and non-governmental organisations are active in defining development and peace-building strategies. Amongst them is the European Union which has become increasingly aware of its political responsibility. Accordingly, the Common Foreign and Security Policy (CFSP) and a Common Security and Defence Policy (CSDP) were formulated to play a more active role in global affairs focussing, inter alia, on being able “to act before a crisis occurs”. With the new EU Global Strategy 2016 this comprises an integrated approach to conflicts and crises acting at all stages of the conflict cycle including resilience-building (European Union 2016). UNEP has postulated the monitoring of natural resource extraction as one of the most important issues of peace building and stabilisation support in conflict regions (UNEP 2009). Earth observation can contribute to more effective monitoring of the exploitation of timber, minerals and other natural resources as well as of developments in and around IDP/refugee camps. A detailed assessment of conflict situations and a better understanding of the situation in a particular region of interest is supported by the integrated analysis of local land cover dynamics complemented by further socio-economic and political information. At EU level earth observation information for the identification, documentation and monitoring of activities in the scope of the aforementioned objectives is delivered through the EU External Action Service supported by the Copernicus programme (formerly GMES - Global Monitoring for Environment and Security). In this context the overall methodological aim of this dissertation is to assess the use of satellite data to detect hotspots of human impact on natural resources that either indicate increasing resource scarcity or can be related to an increasing probability of conflict X intensification by using EO data and advanced remote sensing techniques at different spatial scales. From a socio-political point of view, in order to arrive at a more comprehensive picture of the in-field situation, the remote sensing analysis shall be integrated with statistical information and findings gathered from reports. The strategy of this dissertation is to focus on two case studies, which represent broader application cases, thematically distinct, yet complementary with respect to the overall aim of supporting humanitarian assistance, conflict/crisis prevention as well as post-conflict stabilisation and resilience capacities: (1) population monitoring and environmental impact assessment in the vicinity of IDP/Refugee camps in the Darfur region of Sudan, and (2) monitoring of conflict resources in the Democratic Republic of the Congo. This dissertation with its different elements and case studies demonstrates the capability of earth observation based assessments at different spatial scales for the monitoring of population dynamics caused by migration as well as the detection of conflict-related changes in land cover indicating environmental impact, informal and illicit resource extraction. By introducing a comprehensive multi-scale approach, the advantages of covering large regions while at the same time generating detailed insights into developments within specific areas of interest are demonstrated. Beyond this multi-scale analysis workflow, the overall methodological framework comprises time-series and other pixel-based analyses at subnational and regional scale, OBIA and OBCD approaches as well as GIS concepts and photogrammetric workflows at regional and local scale. In the broader context, the supporting capabilities of geospatial information and earth observation for humanitarian assistance, conflict and crisis prevention as well as postconflict stabilisation and resilience capacities are shown.
View
... In cases where high-density photo- grammetric point clouds are used, ground filtering algorithms have to consider all points as they belong to the ground surface, which makes filtering more challenging. Cooper, Friedman, and Wood (1987), Toutin (2004), Arefi et al. (2009), andTian, Krauss, andReinartz (2014) used stereo-imagery for DTM generation. Sefercik et al. (2013) used the RASCOR software to filter a point cloud derived from stereo WorldView-2 imagery to generate a DTM, which was then compared to a reference DTM and to a lidar-derived DTM. ...
... In cases where high-density photo- grammetric point clouds are used, ground filtering algorithms have to consider all points as they belong to the ground surface, which makes filtering more challenging. Cooper, Friedman, and Wood (1987), Toutin (2004), Arefi et al. (2009), andTian, Krauss, andReinartz (2014) used stereo-imagery for DTM generation. Sefercik et al. (2013) used the RASCOR software to filter a point cloud derived from stereo WorldView-2 imagery to generate a DTM, which was then compared to a reference DTM and to a lidar-derived DTM. ...
Investigating the performances of commercial and non-commercial software for ground filtering of UAV-based point clouds
Article
  • Jan 2018
In recent years, the advent of unmanned aerial vehicles (UAV)-based photogrammetry has enabled the collection of accurate and comprehensive information from the surface of the Earth. Owing to low-altitude flights, it is possible to generate high-density point clouds, which are useful for accurate representation of topography of the land surface. Ground filtering is the removal of the points belonging to above-ground objects in order to retrieve ground points to be used in generating digital terrain models. It is essential in most applications for modelling the environment and is performed by using various types of commercial and non-commercial software. This study investigates the performances of seven widely used ground filtering algorithms on UAV-based point clouds. These algorithms are (1) the adaptive triangulated irregular network implemented into the commercial Agisoft Photoscan Professional software, (2) the multi-scale curvature classification implemented into the commercial global mapper software, (3) the cloth simulation filtering (CSF) applied with a MATLAB script, (4) the interpolation-based Boise Centre Aerospace Laboratory-lidar algorithm embedded in the commercial environment for visualizing images software, (5) the interpolation-based gLiDAR non-commercial software, (6) the 2D progressive morphological algorithms, and (7) elevation threshold with expand window algorithms embedded in the non-commercial airborne lidar data processing and analysis tools software. The results showed that the CSF algorithm presented the best filtering results for both test sites with Total Errors of 6% and 4.5% in the sites 1 and 2, respectively.
View
... Krauss, and Reinartz, 2014). There, the panchromatic image is classified into semantic classes, like buildings, low forest, high forest or ground, and different filtering is applied on DSM regions according to the class. ...
Mapping with Pléiades
Thesis
Full-text available
  • Jun 2019
An end-to-end workflow for mapping with very high resolution satellite data is the pre-requisite for any further semantic analysis. In specific, many applications in remote sensing need the following 3D mapping products: (1) digital surface model, (2) digital terrain model, (3) normalized digital surface model, and (4) ortho-rectified image mosaic. This thesis describes all underlying principles for satellite-based 3D mapping and proposes methods that extract all those products from multi-view stereo satellite imagery in the ground sampling distance of the input data. The study is based on, but not limited to, the Pleiades satellite constellation. Beside an in-depth review of related work, the methodological part proposes solutions for each component of the end-to-end workflow. In particular, this includes optimization of sensor models represented by rational polynomials, epipolar rectification, image matching, spatial point intersection, data fusion, digital terrain model derivation, ortho rectification, and ortho mosaicing. For each step implementation details are proposed and discussed. Another aim of this thesis is a detailed assessment of the resulting output products. Thus, a variety of data sets showing different acquisition scenarios are gathered, allover comprising 24 Pleiades images. First, the accuracies of the 2D and 3D geo-location are analyzed. Second, surface and terrain models are evaluated, including a critical look on the underlying error metrics, and discussing the differences of single stereo, tri-stereo, and multi-view data sets. Overall, 3D accuracies in the range of 0.2 to 0.3 meters in planimetry and 0.2 to 0.4 meters in height are achieved w.r.t. ground control points. Retrieved surface models show normalized median absolute deviations around 0.9 meters in comparison to reference LiDAR data. Multi-view stereo outperforms single stereo in terms of accuracy and completeness of the resulting surface models.
View
... An approach based on iterative geodesic reconstruction tested on Cartosat-1 stereo images is presented in Reference [103]. DTM generation in forested regions based on Cartosat-1 stereo image is reported in Reference [104]. There, the panchromatic image is classified into semantic classes, like buildings, low forest, high forest or ground and different filtering is applied to DSM regions according to the class. ...
Mapping with Pléiades—End-to-End Workflow
Article
Full-text available
  • Sep 2019
In this work, we introduce an end-to-end workflow for very high-resolution satellite-based mapping, building the basis for important 3D mapping products: (1) digital surface model, (2) digital terrain model, (3) normalized digital surface model and (4) ortho-rectified image mosaic. In particular, we describe all underlying principles for satellite-based 3D mapping and propose methods that extract these products from multi-view stereo satellite imagery. Our workflow is demonstrated for the Pléiades satellite constellation, however, the applied building blocks are more general and thus also applicable for different setups. Besides introducing the overall end-to-end workflow, we need also to tackle single building blocks: optimization of sensor models represented by rational polynomials, epipolar rectification, image matching, spatial point intersection, data fusion, digital terrain model derivation, ortho rectification and ortho mosaicing. For each of these steps, extensions to the state-of-the-art are proposed and discussed in detail. In addition, a novel approach for terrain model generation is introduced. The second aim of the study is a detailed assessment of the resulting output products. Thus, a variety of data sets showing different acquisition scenarios are gathered, allover comprising 24 Pléiades images. First, the accuracies of the 2D and 3D geo-location are analyzed. Second, surface and terrain models are evaluated, including a critical look on the underlying error metrics and discussing the differences of single stereo, tri-stereo and multi-view data sets. Overall, 3D accuracies in the range of 0 . 2 to 0 . 3 m in planimetry and 0 . 2 to 0 . 4 m in height are achieved w.r.t. ground control points. Retrieved surface models show normalized median absolute deviations around 0 . 9 m in comparison to reference LiDAR data. Multi-view stereo outperforms single stereo in terms of accuracy and completeness of the resulting surface models.
View
... Arefi et al. (2011) proposed another technique based on iterative geodesic reconstruction and tested it on Cartosat-1 stereo imagery. Tian et al. (2014) generated a DTM for a vegetated area (forest) by classifying a Cartosat-1 stereo image into forest regions and subtracting the relative region heights from the DSM. Beumier and Idrissa (2015) proposed a method consisting of three steps: (1) DSM region segmentation; (2) region selection; and (3) height interpolation. ...
Validating ALOS PRISM DSM-derived surface feature height: Implications for urban volume estimation
Article
Full-text available
  • Dec 2017
Urban volume, such as urban built volume (UBV), can be used as a proxy indicator for measuring the intensity and spatial pattern of urban development, and for characterizing social structure, intensity of economic activity, levels of economic supremacy, and levels of resource consumption. Urban volume estimation requires two basic input data: (1) urban footprint (built footprint for UBV and green footprint for urban green volume (UGV)); and (2) height data for urban features (herein called surface feature height (SFH)). A digital surface model (DSM) and a digital terrain model (DTM) can be used to extract SFH, i.e., by subtracting the DTM from the DSM. Light Detection and Ranging (LiDAR) data are often used to generate DSMs and DTMs. However, the availability of LiDAR data remains limited. The recent release of ALOS World 3D topographic data provides an alternative data source for DSMs and potentially for DTMs. However, the potential of ALOS PRISM DSM for deriving SFH has not been rigorously assessed, especially at the micro level. In this study, we validated six sets of 5 m ALOS PRISM DSM-derived SFH data across six test sites (To-kyo (Japan), Beijing (China), Shanghai (China), Surabaya (Indonesia) , Tsukuba (Japan) , and Lusaka (Zambia)). We described the grid-based method used to derive a DTM from a DSM and how this method was applied. We then validated the derived SFH data through comparison with recorded building height (RBH) data. Across the six test sites, the root-mean-square error (RMSE) of the ALOS PRISM DSM-derived SFH data ranged from 7 m (Tsuku-ba) (approximately 2 building floors) to 81 m (Beijing) (approximately 27 building floors). The ALOS PRISM DSM-derived SFH data for lower buildings (e.g., RBH < 100 m) and smaller and less dense cities (Surabaya, Tsukuba and Lusaka) were more accurate than for high-rise buildings (e.g. RBH > 100 m) and larger and denser cities (Tokyo, Beijing and Shanghai). Factors that may have influenced the validation results were considered, estimation.
View
EVALUATING THE QUALITY OF DIGITAL DATASETS IN TERMS OF AVIATION SAFETY DERIVED FROM STEREO SATELLITE IMAGERY IN SARAWAK, MALAYSIA'S AIRPORTS
Article
Full-text available
  • Mar 2024
High-resolution satellite imagery offers new possibilities in geospatial data acquisition, particularly in 3D city modeling, topographic mapping, and aviation safety. This study evaluates the quality of digital datasets generated from satellite imagery for five airports in Sarawak, as part of the eTOD (Electronic Terrain and Obstacle Data) project conducted by the Civil Aviation Authority of Malaysia (CAAM). The airports under examination include Miri Airport (WBGR), Bintulu Airport (WBGB), Mukah Airport (WBGK), Sibu Airport (WBGS), and Kuching Airport (WBGG). A comprehensive assessment was conducted to evaluate accuracy and reliability by combining stereo satellite imagery with on-site validation. The study yielded digital datasets of exceptionally high accuracy and reliability. The overall error rate remained below 3%, with a mean error value of 0.96 meters across all five airports. Notably, it has been observed that using satellite images as the only data source creates difficulties in detecting Power Transmission Lines and Poles that require precise detection. A multifaceted approach was adopted to increase the detection accuracy, involving site verification measurements, incorporation of external source data (Sarawak Energy Company), and manual interpretation. In conclusion, this study underscores the significant potential of very high-resolution satellite imagery for creating precise and reliable digital datasets for aviation safety. Additionally, the study's emphasis on detecting Power Transmission Lines highlights the need for advanced methods to improve accuracy in challenging scenarios. These findings not only contribute to enhancing aviation safety but also provide valuable insights into the utilization of satellite imagery for similar applications worldwide.
View
View
Filtering of remote sensing point clouds using fuzzy C-means clustering
Article
  • Jan 2020
Standard filtering techniques perform well on digital surface models (DSMs) generated from light detection and ranging (LiDAR) data. However, these techniques have to be tested and evaluated using point clouds obtained by matching of stereo satellite imagery. This paper proposes a new iterative filtering technique based on fuzzy c-means (FCM) clustering. This method is composed of three main steps: (1) a DSM is generated from GeoEye-1 stereo pair imagery, (2) the generated DSM is then reshaped and applied as input data for a FCM clustering process to separate terrain and non-terrain points, and (3) terrain points are then interpolated into a grid digital terrain model (DTM). An urban test area with distinct land use/cover classes covering the north-east part of Cairo city in Egypt has been selected. To evaluate the performance of the proposed method, the filtered DTM was compared against reference data that was generated manually, and type I, type II, and total errors were estimated. Compared with the most commonly used filtering method, progressive triangular irregular network densification (PTD), the proposed approach has identified terrain points much closer to the reference DTM with less variable accuracies. The FCM filter has reduced the mean errors by 2.06, 0.74, and 0.34% for type I, type II, and total errors, respectively. On the other hand, the mean standard deviation (SD) of the differences between the obtained and reference DTMs has been reduced by 0.38 m.
View
View
Assessment of Cartosat-1 and WorldView-2 stereo imagery in combination with a LiDAR-DTM for timber volume estimation in a highly structured forest in Germany
Article
Full-text available
  • Jul 2013
  • FORESTRY
Stereo satellites provide height information of the earth's surface with increasing accuracy. High temporal resolution and wide regional coverage are the great advantages of satellites compared with aerial surveys. There is currently little experience of how accurate forest attributes can be modelled using high-resolution stereo satellite data, especially for highly structured forests in Central Europe. Thus, the potential of Cartosat-1 and WorldView-2 was assessed for timber volume estimation in a complex forest in Germany. Digital surface models were generated using Semi-Global Matching. Canopy height models (CHMs) were computed by subtracting a Light detection and ranging (LiDAR) terrain model. The CHMs were co-registered with field plots of a forest inventory. Explanatory variables were derived from the CHMs for timber volume estimation using regressions. Accuracies were evaluated at plot and stand levels. Results were compared with estimations based on a LiDAR-CHM. At plot level the following root mean squared errors (RMSEs) for timber volume estimation were obtained: 50.26 per cent for Cartosat-1, 44.40 per cent for WorldView-2 and 38.02 per cent for LiDAR. The RMSEs were smaller than the standard deviation of the observed timber volume. The RMSEs at a stand level yielded 21.49 per cent for Cartosat-1, 19.59 per cent for WorldView-2 and 17.14 per cent for LiDAR. The study demonstrates the potential of satellite stereo images for regionalization of sample plot inventories.
View
Generating an optimal DTM from airborne laser scanning data for landslide mapping in a tropical forest environment
Article
Full-text available
  • May 2013
  • GEOMORPHOLOGY
Landslide inventory maps are fundamental for assessing landslide susceptibility, hazard, and risk. In tropical mountainous environments, mapping landslides is difficult as rapid and dense vegetation growth obscures landslides soon after their occurrence. Airborne laser scanning (ALS) data have been used to construct the digital terrain model (DTM) under dense vegetation, but its reliability for landslide recognition in the tropics remains surprisingly unknown. This study evaluates the suitability of ALS for generating an optimal DTM for mapping landslides in the Cameron Highlands, Malaysia. For the bare-earth extraction, we used hierarchical robust filtering algorithm and a parameterization with three sequential filtering steps. After each filtering step, four interpolations techniques were applied, namely: (i) the linear prediction derived from the SCOP++ (SCP), (ii) the inverse distance weighting (IDW), (iii) the natural neighbor (NEN) and (iv) the topo-to-raster (T2R). We assessed the quality of 12 DTMs in two ways: (1) with respect to 448 field-measured terrain heights and (2) based on the interpretability of landslides. The lowest root-mean-square error (RMSE) was 0.89 m across the landscape using three filtering steps and linear prediction as interpolation method. However, we found that a less stringent DTM filtering unveiled more diagnostic micro-morphological features, but also retained some of vegetation. Hence, a combination of filtering steps is required for optimal landslide interpretation, especially in forested mountainous areas. IDW was favored as the interpolation technique because it combined computational times more reasonably without adding artifacts to the DTM than T2R and NEN, which performed relatively well in the first and second filtering steps, respectively. The laser point density and the resulting ground point density after filtering are key parameters for producing a DTM applicable to landslide identification. The results showed that the ALS-derived DTMs allowed mapping and classifying landslides beneath equatorial mountainous forests, leading to a better understanding of hazardous geomorphic problems in tropical regions.
View
DERIVATION OF DIGITAL TERRAIN MODELS IN THE SCOP++ ENVIRONMENT
Article
Full-text available
Airborne laser scanning is widely used for the derivation of terrain information in wooded or open areas but also for the production of building models in cities. For this, the generation of a digital terrain model (DTM) is also required. In this paper the filtering and classification of laser scanner data with iterative robust linear prediction in a hierarchical fashion using data pyramids is described. The coarse-to-fine approach is advantageous because it strengthens the robustness of the method and makes it faster. The results for test data sets of the OEEPE are presented.
View
Evaluation of selected methods for extracting digital terrain models from satellite born digital surface models in urban areas
Conference Paper
Full-text available
  • May 2011
For the generation of 3D city models from satellite stereo imagery beyond the generation of digital surface models (DSM) from stereo data the next crucial step is the separation of urban 3D objects from ground. To do this the most common method is the derivation of a so called digital terrain model (DTM) from the DSM. The DTM should ideally contain only the surface of the ground on which the urban objects are located. Since only the surface of the objects can be seen from space, sophisticated methods have to be developed to gain information of the bare ground. In this paper selected methods for the extraction of a DTM from a DSM are described and evaluated. The evaluation is done by applying the methods to synthetically generated DSMs. These synthetical DSMs are a combination of ground and typical urban objects put on top of it. The application of the DTM extraction methods should recover in turn the original ground model as good as possible. Also the sum of the obtained DTM and the profile of the urban objects should reconstruct the original DSM. The profile of the urban objects ist often referenced as normalized digital elevation model (nDEM). But in general the equation DSM = DTM + nDEM is not always valid – especially for buildings situated on the slope of a hill. If the nDEM would simply be the difference of DSM – DTM the slope of the hill – contained in the DTM – will be reflected on the roof of the buildings. So also an advanced method for derivation of the nDEM from DSM and DTM is presented and tested.
View
Iterative approach for efficient digital terrain model production from CARTOSAT-1 stereo images
Article
Full-text available
  • Jun 2011
  • J APPL REMOTE SENS
This paper proposes a new algorithm for automatic digital terrain model (DTM) generation from high resolution CARTOSAT-1 satellite images. It consists of two major steps: generation of digital surface models (DSM) from stereo scenes and hierarchical image filtering for DTM generation. Automatic georeferencing, dense stereo matching, and interpolation into a regular grid yields a DSM. In the second step, the DSM pixels are classified into ground and nonground regions using an algorithm motivated from gray-scale image reconstruction to suppress unwanted elevated pixels. Nonground regions, i.e., 3D objects as well as outliers are iteratively separated from the ground regions. The generated DTM is qualitatively and quantitatively evaluated. Height profiles and comparisons between the generated DSM, derived DTM, and ground truth data are presented. The evaluation indicates that almost all nonground objects regardless of their size are eliminated and appropriate results are archived in hilly as well as smooth residential areas.
View
View
Advanced DTM generation from LiDAR data
Article
Full-text available
  • Oct 2011
The introduction of laser scanning has triggered off a revolution in topographic terrain capturing, especially in the generation of digital terrain models (DTM). In this article refined methods for the restitution of airborne LIDAR data are presented which have been developed at the Institute of Photogrammetry and Remote Sensing (Institut für Photogrammetrie und Fernerkundung, I.P.F.) at Vienna University of Technology. First, a technique for the calibration of laser scanner data is introduced. The (height) discrepancies between overlapping strips, as well as control points with known co-ordinates are utilised for a simultaneous adjustment and transformation of all strips into a state wide co-ordinate system. The next step of LIDAR data processing are the filtering (elimination of vegetation and building points, generally off-terrain points) and the interpolation of the (bald earth) surface. The method, developed at the I.P.F., distinguishes itself in the integration of filtering and terrain interpolation in one process (advantage: even in steep terrain ground points are classified correctly) as well as in the application of data pyramids (advantage: even in very dense forest areas and on large buildings, off-terrain points are eliminated). In order to generate a terrain model with high geo-morphological quality, methods are required for deriving structural line information (e.g. break lines) from laser scanner data. The first method which will be presented, proceeds by a simulation of rain fall over the preliminary DTM (water flow analysis). This yields an identification of the pits with their pit base and the outflow (overflow) point.
View
Random Forest
Article
  • Jan 2001
View
Region-based automatic building and forest change detection on Cartosat-1 stereo imagery
Article
  • May 2013
  • ISPRS J PHOTOGRAMM
In this paper a novel region-based method is proposed for change detection using space borne panchromatic Cartosat-1 stereo imagery. In the first step, Digital Surface Models (DSMs) from two dates are generated by semi-global matching. The geometric lateral resolution of the DSMs is 5 m × 5 m and the height accuracy is in the range of approximately 3 m (RMSE). In the second step, mean-shift segmentation is applied on the orthorectified images of two dates to obtain initial regions. A region intersection following a merging strategy is proposed to get minimum change regions and multi-level change vectors are extracted for these regions. Finally change detection is achieved by combining these features with weighted change vector analysis. The result evaluations demonstrate that the applied DSM generation method is well suited for Cartosat-1 imagery, and the extracted height values can largely improve the change detection accuracy, moreover it is shown that the proposed change detection method can be used robustly for both forest and industrial areas.
View
A comparison between photogrammetry and laser scanning. ISPRS J Photogramm Remote Sens 54(2/3):83-94
Article
  • Jul 1999
  • ISPRS J PHOTOGRAMM
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.
View