ArticlePDF Available

Exploring the feasibility of unmanned aerial vehicles and thermal imaging for ungulate surveys in forests - preliminary results

Authors:

Abstract and Figures

Effective wildlife management and conservation require reliable assessments of animal abundance. However, no ungulate monitoring methods is entirely satisfying in terms of cost-effectiveness and accuracy. A new method combining unmanned aerial vehicles (drones) and thermal infrared (TIR) imaging may have great potential as a tool for ungulate surveys. Drones enable safe operations at low flying altitudes, and at night – a time that often offers the optimal conditions for wildlife monitoring. To assess the feasibility of the proposed method we used fixed-wing drones with TIR cameras to conduct test surveys in Drawieński National Park, Poland. We demonstrated that ungulate thermal signatures are visible both in leafless deciduous and in pine-dominated coniferous forests. Survey timing highly influenced the results – the best quality thermal images were obtained at sunrise, late evening, and at night. Our preliminary results indicated that thermal surveys from drones are a promising method for ungulate enumeration. We demonstrated that with ground resolution of ~10 cm it is possible to visibly distinguish large species (i.e. red deer) and achieve a good level of area coverage. The main challenges of the method are difficulties in species identification due to relatively low resolution of TIR cameras, regulations limiting drone operations to visual line of sight, and high dependence on weather.
Content may be subject to copyright.
A preview of the PDF is not available
... Although the area covered by the quadcopter (16.2 km 2 ) is on the higher end of what other studies counting animals with UAV have reported [54], the low density of Svalbard reindeer still makes it challenging to obtain a large enough sample size in relation to the area covered by the UAV. Doing repeat surveys over the same transect lines would increase precision by increasing sample size and is recommended for low-density animals [55]. Note that this may still lead to biased estimates if the area is small and does not cover all habitat characteristics [55]. ...
... Doing repeat surveys over the same transect lines would increase precision by increasing sample size and is recommended for low-density animals [55]. Note that this may still lead to biased estimates if the area is small and does not cover all habitat characteristics [55]. For this wide-ranging species, unbiased estimates require large enough areas to capture the density gradients across the vegetation. ...
Preprint
Conservation of wildlife depends on precise and unbiased knowledge on the abundance and distribution of species. It is challenging to choose appropriate methods to obtain a sufficiently high detectability and spatial coverage matching the species characteristics and spatiotemporal use of the landscape. In remote regions, such as in the Arctic, monitoring efforts are often resource-intensive and there is a need for cheap and precise alternative methods. Here, we compare an uncrewed aerial vehicle (UAV; quadcopter) pilot-survey of the non-gregarious Svalbard reindeer to traditional population abundance surveys from ground and helicopter to investigate whether UAVs can be an efficient alternative technology. We found that the UAV survey underestimated reindeer abundance compared to the traditional abundance surveys when used at management relevant spatial scales. Observer variation in reindeer detection on UAV imagery was influenced by the RGB greenness index and mean blue channel. In future studies, we suggest to test long-range fixed-wing UAVs to increase the sample size of reindeer and area coverage and incorporate detection probability in animal density models from UAV imagery. In addition, we encourage focus on more efficient post-processing techniques, including automatic animal object identification with machine learning and analytical methods that account for uncertainties.
... Thermal infrared cameras have also proven successful in counting fruit bats such as flying-foxes (Pteropus spp.), with estimates of colony size 2.05 and 1.92 times higher than ground-based counts by the experimenter and Australian government counters respectively, providing more accurate and precise population density counts (McCarthy et al. 2021(McCarthy et al. , 2022. In terms of densely forested habitats, drones mounted with TIR cameras have been used to successfully survey ungulates including red deer (Cervus elaphus), Javan deer (Rusa timorensis) and water buffalo (Bubalus arnee) in coniferous and deciduous forests of Poland (Witczuk et al. 2017), and also wildlife in the tropics (Rahman and Rahman 2021). Rahman and Rahman (2021) surveyed an Indonesian forested-urban area (40% forested) to compare TIR with ground-based camera traps and transects. ...
... This would also help to tease apart the effects of the physical and behavioral, potentially co-occurring effects. Additionally, improved resolution of TIR cameras could allow for more accurate, species level, identifications (Burke et al. 2019b;Witczuk et al. 2017). ...
Article
Full-text available
The payload size and commercial availability of thermal infrared cameras mounted on drones has initiated a new wave in the potential for conservationists and researchers to survey , count and detect wildlife, even the most complex of habitats such as forest canopies. However, several fundamental design and methodological questions remain to be tested before standardized monitoring approaches can be broadly adopted. We test the impact of both the speed of drone flights and diel flight period on tropical rainforest canopy wildlife detections. Detection and identification rates differ between both flight speeds and diel time. Overall ~ 36% more detections were made during slower flight speeds, along with a greater ability to categorize taxonomic groups. Flights conducted at 3am resulted in ~ 67% more detections compared to flights conducted at 7am (the diel period with the lowest detection rate). However, 112% more detections could be identified to taxonomic group in 7am flights compared with 3am flights-due to the types of wildlife being identified and the assistance of the RGB camera. Although, this technology holds great promise for carrying out surveys in structurally complex and poorly known ecosystems like forest canopies, there is more to do in further methodological testing, and building automated post-processing systems. Our results suggest that drone studies in the same habitat types, with the same animal densities, could be off by multiples if flown during different times and/or at different speeds. The difference could be an alarming 5-6x variation in animal detections or identification depending on changes in these two factors alone.
... Therefore, the counts are often substituted with different type of indices [moose seen per hunter; 26,camera trap and dung; 27,28]. However, only recently UAVs are applied in surveys of ungulates, where restricted flight range relative to the spatial scale of interest for management and difficulties detecting and identifying deer, have been major obstacles [3,8,[29][30][31][32][33]. ...
... This is likely attributed to the low spatial extent of the UAV area, which was approximately half of the ground DS sampling area. Doing repeat surveys over the same transect lines may increase accuracy [30]. The helicopter survey had a higher density (5.4 reindeer / km 2 ) than the UAV survey and ground DS survey (1.4-2.4 reindeer / km 2 ). ...
Preprint
Full-text available
Conservation of wildlife depends on precise and unbiased knowledge on the abundance and distribution of species. A challenge is to choose appropriate methods to obtain a sufficiently high detectability and spatial coverage matching the species characteristics and spatiotemporal use of the landscape. In remote areas, such as in the Arctic, monitoring efforts are often resource demanding and there is a need for cheap and precise alternative methods. Here, we compare an UAV pilot-survey to traditional population abundance surveys from ground and helicopter of the non-gregarious Svalbard reindeer to investigate whether small quadcopter UAVs can be an efficient alternative technology. We find that estimates of reindeer abundances from UAV imagery have lower precision and are more time consuming than present abundance surveys when used at management relevant spatial scales. We suggest that more efficient long-range fixed-wing UAVs should be evaluated for the job to increase the sampled area by UAV. In addition, the method will depend on the availability of more efficient post-processing methods including automatic animal object identification with machine learning and analytical methods that account for uncertainties.
... Most publications on the use of thermal imaging to facilitate wildlife surveys relate to its use with unmanned aerial vehicles (UAVs) to count large mammals (Lhoest et al 2015, Witczuk et al 2018, Beaver et al 2020 and to locate birds' nests, either on the ground (Israel & Reinhard 2017, Santangeli et al 2020, Shewring & Vafidis 2021 or over water (Bushaw et al 2020). The advantages of this approach are that they increase both survey efficacy and human safety when compared to manned aerial surveys (Beaver et al 2020). ...
Article
The West Midlands Ringing Group (WMRG) have been using thermal-imaging equipment since 2016 to locate elusive birds for ringing by night and by day, to survey roosting birds in reedbeds, to monitor nets at night and to check nest boxes for occupancy. The technique has not only transformed the catch rate for ringing surveys of birds such as Skylark Alauda arvensis, Woodcock Scolopax rusticola, Snipe Gallinago gallinago and Jack Snipe Lymnocryptes minimus but also resulted in an increase in survey efficiency, a reduction in disturbance of birds and an improvement in the accuracy of counts for birds of conservation concern. Here we present case studies from WMRG’s work and an overview of the advantages, constraints and potential applications for the use of thermal imaging in ornithological surveys.
Article
Full-text available
The Miombo woodland is the most extensive tropical woodland in south-central Africa. However, field sample plot data on forest cover changes, species distribution and carbon stocks in the Miombo ecoregion are inadequate for effective forest management. Owing to logistical challenges that come with field-based inventory methods, remote sensing plays an important role in supplementing field methods to fill in data gaps. Traditional satellite and manned aircraft remote sensing platforms have their own advantages and limitations. The advent of unmanned aerial systems (UASs) has made it possible to acquire forest data at unprecedented spatial and temporal scales. UASs are adaptable to various forest applications in terms of providing flexibility in data acquisition with different sensors (RGB, multispectral, hyperspectral, thermal and light detection and ranging (lidar)) at a convenient time. To highlight possible applications in the Miombo woodlands, we first provide an overview of the Miombo woodlands and recent progress in remote sensing with small UASs. An overview of some potential forest applications was undertaken to identify key prospects and challenges for UAS applications in the Miombo region, which will provide expertise and guidance upon which future applications in the Miombo woodlands should be based. While much of the potential of using UASs for forest data acquisition in the Miombo woodlands remains to be realized, it is likely that the next few years will see such systems being used to provide data for an ever-increasing range of forest applications.
Article
Over the last decade, Multi-UAS system (MUS) have found a variety of applications such as surveillance, data acquisition, search and rescue, and delivery tasks in various environments. However, for reliable operation of MUS, it is critical to provide safety guarantee conditions assuring obstacle collision avoidance, as well as, inter-UAS collision avoidance. In this work, we apply the principles of continuum mechanics to develop a physics-inspired algorithm to ascertain safe and resilient operation of a MUS in the presence of disturbances and unforeseen UAS failure(s). In particular, the proposed approach consists of two modes: (i) Homogeneous Deformation Mode (HDM), and (ii) Failure Resilient Mode (FRM). We formally specify transitions between HDM and FRM using Cooperative Localization (CL) approach to quantify UAS tracking error and detect anomalous conditions due to UAS failure.
Preprint
Full-text available
Drones equipped with thermal cameras have recently become readily available, broadening the possibilities for monitoring wildlife. The European hare (Lepus europaeus) is a nocturnal mammal that is closely monitored in Denmark due to populations declining since the mid-1900s. The limitations of current population assessment methods, such as spotlight counts and hunting game statistics, could be overcome by relying on drone surveys with thermal imaging for population counts. The aim of this study was to investigate the use of a DJI Mavic 2 Enterprise Advanced drone with thermal imaging as a tool for monitoring the Danish hare population. Multiple test flights were conducted over agricultural areas in Denmark in spring 2022, testing various flight altitudes, camera settings, and recording methods. The test flights were used to suggest a method for identifying and counting hare. The applied use of this methodology was then evaluated through a case survey that had the aim of identifying and counting hare over an agricultural area of 242 ha. Hare could be detected with thermal imaging at flight altitudes up to 80 m and it was possible to fly as low as 40 m without observing direct behavior changes. Thermal images taken at these altitudes also provided enough detail to differentiate between species and animal body size proved to be a good species indicator. The case study confirmed the use of thermal imaging based drone surveys to identify hare and conduct population counts, thus, indicating the suggested methodology as a viable alternative to traditional counting methods.
Article
In the field of species conservation, the use of unmanned aerial vehicles (UAV) is increasing in popularity as wildlife observation and monitoring tools. With large datasets created by UAV-based species surveying, the need arose to automate the detection process of the species. Although the use of computer learning algorithms for wildlife detection from UAV-derived imagery is an increasing trend, it depends on a large amount of imagery of the species to train the object detector effectively. However, there are alternatives like object-based image analysis (OBIA) software available if a large amount of imagery of the species is not available to develop a computer-learned object detector. The study tested the semi-automated detection of reintroduced Arabian Oryx (O. leucoryx), using the specie's coat sRGB-colour profiles as input for OBIA to identify adult O. leucoryx, applied to UAV acquired imagery. Our method uses lab-measured spectral reflection of hair sample values, collected from captive O. leucoryx as an input for OBIA ruleset to identify adult O. leucoryx from UAV survey imagery using semi-automated supervised classification. The converted mean CIE Lab reflective spectrometry colour values of n = 50 hair samples of adult O. leucoryx to 8-bit sRGB-colour profiles of the species resulted in the red-band value of 157.450, the green-band value of 151.390 and blue-band value of 140.832. The sRGB values and a minimum size permitter were added as the input of the OBIA ruleset identified adult O. leucoryx with a high degree of efficiency when applied to three UAV census datasets. Using species sRGB-colour profiles to identify re-introduced O. leucoryx and extract location data using a non-invasive UAV-based tool is a novel method with enormous application possibilities. Coat refection sRGB-colour profiles can be developed for a range of species and customised to autodetect and classify the species from remote sensing data.
Article
Full-text available
Estimating animal populations is critical for wildlife management. Aerial surveys are used for generating population estimates, but can be hampered by cost, logistical complexity, and human risk. Additionally, human counts of organisms in aerial imagery can be tedious and subjective. Automated approaches show promise, but can be constrained by long setup times and difficulty discriminating animals in aggregations. We combine unmanned aircraft systems (UAS), thermal imagery and computer vision to improve traditional wildlife survey methods. During spring 2015, we flew fixed-wing UAS equipped with thermal sensors, imaging two grey seal (Halichoerus grypus) breeding colonies in eastern Canada. Human analysts counted and classified individual seals in imagery manually. Concurrently, an automated classification and detection algorithm discriminated seals based upon temperature, size, and shape of thermal signatures. Automated counts were within 95–98% of human estimates; at Saddle Island, the model estimated 894 seals compared to analyst counts of 913, and at Hay Island estimated 2188 seals compared to analysts’ 2311. The algorithm improves upon shortcomings of computer vision by effectively recognizing seals in aggregations while keeping model setup time minimal. Our study illustrates how UAS, thermal imagery, and automated detection can be combined to efficiently collect population data critical to wildlife management.
Article
Full-text available
Tropical deforestation continues to be a major driver of biodiversity loss and greenhouse gas emissions. Remote sensing technology is increasingly used to assess changes in forest cover, species distributions and carbon stocks. However, satellite and airborne sensors can be prohibitively costly and inaccessible for researchers in developing countries. Here, we describe the development and use of an inexpensive (<$2,000) unmanned aerial vehicle for surveying and mapping forests and biodiversity (referred to as ‘Conservation Drone’ hereafter). Our prototype drone is able to fly pre-programmed missions autonomously for a total flight time of ~25 minutes and over a distance of ~15 km. Non-technical operators can program each mission by defining waypoints along a flight path using an open-source software. This drone can record videos at up to 1080 pixel resolution (high definition), and acquire aerial photographs of <10 cm pixel resolution. Aerial photographs can be stitched together to produce real-time geo-referenced land use/cover maps of surveyed areas. We evaluate the performance of this prototype Conservation Drone based on a series of test flights in Aras Napal, Sumatra, Indonesia. We discuss the further development of Conservation Drone 2.0, which will have a bigger payload and longer range. Initial tests suggest a flight time of ~50 minutes and a range of ~25 km. Finally, we highlight the potential of this system for environmental and conservation applications, which include near real-time mapping of local land cover, monitoring of illegal forest activities, and surveying of large animal species.
Article
Full-text available
Assessing temporal changes in abundance indices is an important issue in the management of large herbivore populations. The drive counts method has been frequently used as a deer abundance index in mountainous regions. However, despite an inherent risk for observation errors in drive counts, which increase with deer density, evaluations of the utility of drive counts at a high deer density remain scarce. We compared the drive counts and mark-resight (MR) methods in the evaluation of a highly dense sika deer population (MR estimates ranged between 11 and 53 individuals/km²) on Nakanoshima Island, Hokkaido, Japan, between 1999 and 2006. This deer population experienced two large reductions in density; approximately 200 animals in total were taken from the population through a large-scale population removal and a separate winter mass mortality event. Although the drive counts tracked temporal changes in deer abundance on the island, they overestimated the counts for all years in comparison to the MR method. Increased overestimation in drive count estimates after the winter mass mortality event may be due to a double count derived from increased deer movement and recovery of body condition secondary to the mitigation of density-dependent food limitations. Drive counts are unreliable because they are affected by unfavorable factors such as bad weather, and they are cost-prohibitive to repeat, which precludes the calculation of confidence intervals. Therefore, the use of drive counts to infer the deer abundance needs to be reconsidered.
Article
Full-text available
Unmanned aircraft systems (UAS) - also called unmanned aerial vehicles (UAVs) or drones - are an emerging tool that may provide a safer, more cost-effective, and quieter alternative to traditional research methods. We review examples where UAS have been used to document wildlife abundance, behavior, and habitat, and illustrate the strengths and weaknesses of this technology with two case studies. We summarize research on behavioral responses of wildlife to UAS, and discuss the need to understand how recreational and commercial applications of this technology could disturb certain species. Currently, the widespread implementation of UAS by scientists is limited by flight range, regulatory frameworks, and a lack of validation. UAS are most effective when used to examine smaller areas close to their launch sites, whereas manned aircraft are recommended for surveying greater distances. The growing demand for UAS in research and industry is driving rapid regulatory and technological progress, which in turn will make them more accessible and effective as analytical tools.
Article
Full-text available
Unmanned aerial vehicles (UAVs) represent a new frontier in environmental research. Their use has the potential to revolutionise the field if they prove capable of improving data quality or the ease with which data are collected beyond traditional methods. We apply UAV technology to wildlife monitoring in tropical and polar environments and demonstrate that UAV-derived counts of colony nesting birds are an order of magnitude more precise than traditional ground counts. The increased count precision afforded by UAVs, along with their ability to survey hard-to-reach populations and places, will likely drive many wildlife monitoring projects that rely on population counts to transition from traditional methods to UAV technology. Careful consideration will be required to ensure the coherence of historic data sets with new UAV-derived data and we propose a method for determining the number of duplicated (concurrent UAV and ground counts) sampling points needed to achieve data compatibility.
Article
High-resolution aerial images allow detailed analyses of periglacial landforms, which is of particular importance in light of climate change and resulting changes in active layer thickness. The aim of this study is to show possibilities of using UAV-based photography to perform spatial analysis of periglacial landforms on the Demay Point peninsula, King George Island, and hence to supplement previous geomorphological studies of the South Shetland Islands.Photogrammetric flights were performed using a PW-ZOOM fixed-winged unmanned aircraft vehicle. Digital elevation models (DEM) and maps of slope and contour lines were prepared in ESRI ArcGIS 10.3 with the Spatial Analyst extension, and three-dimensional visualizations in ESRI ArcScene 10.3 software.Careful interpretation of orthophoto and DEM, allowed us to vectorize polygons of landforms, such as (i) solifluction landforms (solifluction sheets, tongues, and lobes); (ii) scarps, taluses, and a protalus rampart; (iii) patterned ground (hummocks, sorted circles, stripes, nets and labyrinths, and nonsorted nets and stripes); (iv) coastal landforms (cliffs and beaches); (v) landslides and mud flows; and (vi) stone fields and bedrock outcrops.We conclude that geomorphological studies based on commonly accessible aerial and satellite images can underestimate the spatial extent of periglacial landforms and result in incomplete inventories. The PW-ZOOM UAV is well suited to gather detailed geomorphological data and can be used in spatial analysis of periglacial landforms in the Western Antarctic Peninsula region.
Article
In this paper we describe an unmanned aerial system equipped with a thermal-infrared camera and software pipeline that we have developed to monitor animal populations for conservation purposes. Taking a multi-disciplinary approach to tackle this problem, we use freely available astronomical source detection software and the associated expertise of astronomers, to efficiently and reliably detect humans and animals in aerial thermal-infrared footage. Combining this astronomical detection software with existing machine learning algorithms into a single, automated, end-to-end pipeline, we test the software using aerial video footage taken in a controlled, field-like environment. We demonstrate that the pipeline works reliably and describe how it can be used to estimate the completeness of different observational datasets to objects of a given type as a function of height, observing conditions etc. -- a crucial step in converting video footage to scientifically useful information such as the spatial distribution and density of different animal species. Finally, having demonstrated the potential utility of the system, we describe the steps we are taking to adapt the system for work in the field, in particular systematic monitoring of endangered species at National Parks around the world.
Article
Wildlife management is based on various measurements representative of the health of populations and their habitats. Some agencies are focusing on animal surveys to manage species such as white-tailed deer (Odocoileus virginianus). Current survey methods are faced with the challenge of reduced operating costs as well as estimating and correcting detection biases. Our pilot study (data collected on 6 Nov 2012 at Saint-David-de-Falardeau, QC, Canada) assessed the potential of a new approach detect and count deer based on visible and thermal infrared image processing at very-high spatial resolutions using an unmanned aerial system (UAS). Supervised and unsupervised pixel-based image classification approaches as well as object-based image analysis (OBIA) were assessed for different spatial resolutions and with different combinations of spectral bands. None of the pixel-based approaches were effective for detecting deer. The OBIA approach detected deer with a rate of up to 100% under the best conditions by using a combination of visible and thermal infrared imagery at a spatial resolution of 0.8 cm/pixel. Overall, this approach had an average detection rate of 0.5, which is comparable to conventional aerial surveys. Visual obstructions by coniferous canopy and the spectral confusion associated with certain elements (e.g., bare soil, rocks) are problems that remain unresolved. Using UASs with image processing for surveys of deer and other species of large mammals is promising, but currently limited by the flight range of unmanned aerial vehicles and the associated regulations.