[en] As the need to accurately monitor key-species populations grows amid increasing pressures on global biodiversity, the counting of large mammals in savannas has traditionally relied on the Systematic-Reconnaissance-Flight (SRF) technique using light aircrafts and human observers. However, this method has limitations, including non-systematic human errors. In recent years, the Oblique-Camera-Count (OCC) approach developed in East Africa has utilized cameras to capture high-resolution imagery replicating aircraft observers’ oblique view. Whilst demonstrating that human observers have missed many animals, OCC relies on labor-intensive human interpretation of thousands of images. This study explores the potential of Deep Learning (DL) to reduce the interpretation workload associated with OCC surveys. Using oblique aerial imagery of 2.1 hectares footprint collected during an SRF-OCC survey of Queen Elizabeth Protected Area in Uganda, a DL model (HerdNet) was trained and evaluated to detect and count 12 wildlife and livestock mammal species. The model’s performance was assessed both at the animal instance-based and image-based levels, achieving accurate detection performance (F1 score of 85%) in positive images (i.e. containing animals) and reducing manual interpretation workload by 74% on a realistic dataset showing less than 10% of positive images. However, it struggled to differentiate visually related species and overestimated animal counts due to false positives generated by landscape items resembling animals. These challenges may be addressed through improved training and verification processes. The results highlight DL’s potential to semi-automate processing of aerial survey wildlife imagery, reducing manual interpretation burden. By incorporating DL models into existing counting standards, future surveys may increase sampling efforts, improve accuracy, and enhance aerial survey safety.
FRIA - Fonds pour la Formation à la Recherche dans l'Industrie et dans l'Agriculture [BE] Global Conservation of California [US-CA]
Funding text :
The annotation work was supported under a grant from Global Conservation of California, USA to R. Lamprey and the WildSpace-Image-Analytics team of Uganda (www.wildspace-image-analytics.com). Sharing of annotation data was conducted under a Memorandum of Understanding between Uganda Conservation Foundation (UCF) and the University of Liege. The original aerial 2018 survey of QENP, from which this experimental imagery was collected, was funded by UCF with support from Global Conservation, Vulcan Inc., Save the Elephants and the Uganda Wildlife Authority. We are grateful to Jeff Morgan of Global Conservation and Mike Keigwin of the Uganda Conservation Foundation who supported R. Lamprey and the WildSpace-Image-Analytics team in Uganda in conducting the first-stage image annotation work of this project. We would like to thank the Uganda Wildlife Authority for their assistance in the original QEPA aerial survey of 2018, and especially to Mr Charles Tumwesigye, Director of Conservation at UWA, who kindly obtained the necessary authorizations.
Bröker K. C. A. Hansen R. G. Leonard K. E. Koski W. R. Heide-Jørgensen M. P. (2019). A comparison of image and observer based aerial surveys of narwhal. Mar. Mammal Sci. 35 (4), 1253–1279. doi: 10.1111/mms.12586
Brummitt N. Regan E. C. Weatherdon L. V. Martin C. S. Geijzendorffer I. R. Rocchini D. et al. (2017). Taking stock of nature: Essential biodiversity variables explained. Biol. Conserv. 213, 252–255. doi: 10.1016/j.biocon.2016.09.006
Buckland S. T. Anderson D. R. Burnham K. P. Laake J. L. Borchers D. L. Thomas L. (2004). Advanced distance sampling. (Oxford, United Kingdom: Oxford University Press).
Caughley G. (1974). Bias in aerial survey. J. Wildl. Manage. 38 (4), 921. doi: 10.2307/3800067
Caughley G. (1977). Sampling in aerial survey. J. Wildl. Manage. 41 (4), 605. doi: 10.2307/3799980
Caughley G. Goddard J. (1975). Abundance and distribution of elephants in the Luangwa Valley, Zambia. Afr. J. Ecol. 13 (1), 39–48. doi: 10.1111/j.1365-2028.1975.tb00122.x
CITES-MIKE (2019). Monitoring the Illegal Killing of Elephants: Aerial Survey Standards for the MIKE Programme. Version 3.0. Convention on International Trade in Endangered Species - Monitoring the Illegal Killing of Elephants Programme (CITES-MIKE) (Nairobi, Kenya: United Nations Environment Programme).
Craig G. C. (2012). Aerial survey standards for the CITES-MIKE Programme, Version 2 (UNEP/DELC, Nairobi, Kenya: CITES-MIKE Programme).
Delplanque A. Foucher S. Lejeune P. Linchant J. Théau J. (2021). Multispecies detection and identification of African mammals in aerial imagery using convolutional neural networks. In: Remote Sens. Ecol. Conserv. 8 (2), pp.166–pp.179. doi: 10.1002/rse2.234
Delplanque A. Foucher S. Lejeune P. Théau J. (2022). “Counting African mammal herds in aerial imagery using deep learning: are anchor-based algorithms the most suitable?,” in 10th International Conference on Agro-Geoinformatics and 43rd Canadian Symposium on Remote Sensing, Québec, Canada. 14 July 2022. https://hdl.handle.net/2268/293320.
Delplanque A. Foucher S. Théau J. Bussière E. Vermeulen C. Lejeune P. (2023). From crowd to herd counting: how to precisely detect and count African mammals using aerial imagery and deep learning? ISPRS J. Photogrammetry Remote Sens. 197, 167–180. doi: 10.1016/j.isprsjprs.2023.01.025
Dutta A. Zisserman A. (2019). “The VIA annotation software for images, audio and video,” in Proceedings of the 27th ACM International Conference on Multimedia, New York, USA. (Association for Computing Machinery, New York (United States of America)), p.2276–2279. doi: 10.1145/3343031.3350535
Eberhardt L. L. (1978). Transect methods for population studies. J. Wildl. Manage. 42 (1), 1–31. doi: 10.2307/3800685
Eikelboom J. A. J. Wind J. van de Ven E. Kenana L. M. Schroder B. de Knegt H. J. et al. (2019). Improving the precision and accuracy of animal population estimates with aerial image object detection. Methods Ecol. Evol. 10 (11), 1875–1887. doi: 10.1111/2041-210x.13277
Fleming P. J. S. Tracey J. P. (2008). Some human, aircraft and animal factors affecting aerial surveys: how to enumerate animals from the air. Wildl. Res. 35 (4), 258. doi: 10.1071/wr07081
Grimsdell J. J. R. Westley S. B. (1981). Low-level aerial survey techniques. Addis Ababa: Monograph 4 (Addis Ababa (Ethiopia): International Livestock Centre for Africa).
Gwynne M. D. Croze H. (1975). “East African habitat monitoring practice: a review of methods and application,” in Evaluation and Mapping of Tropical African Rangelands; Proceedings of the Seminar in Bamako, Mali. (International Livestock Centre for Africa, Addis Ababa (Ethiopia)), 95–136.
Jachmann H. (2001). Estimating the abundance of African wildlife (New York United States of America: Springer Science & Business Media).
Jachmann H. (2002). Comparison of aerial counts with ground counts for large African herbivores. J. Appl. Ecol. 39 (5), 841–852. doi: 10.1046/j.1365-2664.2002.00752.x
Jetz W. McGeoch M. A. Guralnick R. Ferrier S. Beck J. Costello M. J. et al. (2019). Essential biodiversity variables for mapping and monitoring species populations. Nat. Ecol. Evol. 3 (4), 539–551. doi: 10.1038/s41559-019-0826-1
Jolly G. M. (1969). Sampling methods for aerial censuses of wildlife populations. East African. Agric. Forest. J. 34 (1), 46–49. doi: 10.1080/00128325.1969.11662347
Kellenberger B. Marcos D. Tuia D. (2018). Detecting mammals in UAV images: Best practices to address a substantially imbalanced dataset with deep learning. Remote Sens. Environ. 216, 139–153. doi: 10.1016/j.rse.2018.06.028
Kruger J. M. Reilly B. K. Whyte I. J. (2008). Application of distance sampling to estimate population densities of large herbivores in Kruger National Park. Wildl. Res. 35 (4), 371–376. doi: 10.1071/WR07084
Lamprey R. H. Keigwin M. Tumwesigye C. (2023)A high-resolution aerial camera survey of Uganda’s Queen Elizabeth Protected Area improves detection of wildlife and delivers a surprisingly high estimate of the elephant population (Accessed February 15, 2023).
Lamprey R. H. Ochanda D. Brett R. Tumwesigye C. Douglas-Hamilton I. (2020). Cameras replace human observers in multi-species aerial counts in Murchison Falls, Uganda. Remote Sens. Ecol. Conserv. 6, 529–545. doi: 10.1002/rse2.154
Lamprey R. H. Pope F. Ngene S. Norton-Griffiths M. Frederick H. Okita-Ouma B. et al. (2019). Comparing an automated high-definition oblique camera system to rear-seat-observers in a wildlife survey in Tsavo, Kenya: Taking multi-species aerial counts to the next level. Biol. Conserv. 241, 108243. doi: 10.1016/j.biocon.2019.108243
LeCun Y. Bengio Y. Hinton G. (2015). Deep learning. Nature 521 (753), 436–444. doi: 10.1038/nature14539
Lethbridge M. Stead M. Wells C. (2019). Estimating kangaroo density by aerial survey: A comparison of thermal cameras with human observers. Wildl. Res. 46 (8), 639–648. doi: 10.1071/WR18122
Naudé J. Joubert D. (2019). “The aerial elephant dataset: A new public benchmark for aerial object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, (Institute for Electrical and Electronics Engineers (IEEE), New York City (United States of America)) Vol. pp. 48–55.
Norton-Griffiths M. (1978). Counting animals (Washington DC: African Wildlife Foundation Handbook Number 1).
Ogutu J. O. Piepho H. P. Said M. Y. Ojwang G. O. Njino L. W. Kifugo S. C. et al. (2016). Extreme wildlife declines and concurrent increase in livestock numbers in Kenya: What are the causes? PloS One 11 (9), 1–46. doi: 10.1371/journal.pone.0163249
Oksuz K. Cam B. C. Kalkan S. Akbas E. (2020). Imbalance problems in object detection: A review. IEEE Trans. Pattern Anal. Mach. Intell. 43 (10), 3388–3415. doi: 10.1109/TPAMI.2020.2981890
Ottichilo W. K. Khaemba W. M. (2001). Validation of observer and aircraft calibration for aerial surveys of animals. Afr. J. Ecol. 39 (1), 45–50. doi: 10.1046/j.1365-2028.2001.00268.x
PAEAS (2014). Aerial survey standards and guidelines for the Pan-African Elephant Aerial Survey (Seattle, USA: Vulcan Inc).
Peng J. Wang D. Liao X. Shao Q. Sun Z. Yue H. et al. (2020). Wild animal survey using UAS imagery and deep learning: modified Faster R-CNN for kiang detection in Tibetan Plateau. ISPRS J. Photogrammetry Remote Sens. 169, 364–376. doi: 10.1016/j.isprsjprs.2020.08.026
Pennycuick C. J. Western D. (1969). An investigation of some sources of bias in aerial transect sampling of large mammal populations. Afr. J. Ecol. 10 (3), pp.175–pp.191. doi: 10.1111/j.1365-2028.1972.tb00726.x
Sarwar F. Griffin A. Rehman S. U. Pasang T. (2021). Detecting sheep in UAV images. Comput. Electron. Agric. 187, 106219. doi: 10.1016/j.compag.2021.106219
Schlossberg S. Chase M. J. Griffin C. R. (2016). Testing the accuracy of aerial surveys for large mammals: An experiment with African savanna elephants (Loxodonta africana). PloS One 11 (10), e0164904. doi: 10.1371/journal.pone.0164904
Shepley A. Falzon G. Meek P. Kwan P. (2021). Automated location invariant animal detection in camera trap images using publicly available data sources. Ecol. Evol. 11, 4494–4506. doi: 10.1002/ece3.7344
Stelfox J. G. Peden D. G. (1981). “The aerial survey programme of the Kenya Rangeland Ecological Monitoring Unit,” in Low-Level Aerial Survey Techniques. Eds. Grimsdell J. J. R. Westley S. B. (Addis Ababa, Ethiopia: ILCA Monographs 4), 69–83.
Tabak M. A. Norouzzadeh M. S. Wolfson D. W. Sweeney S. J. VerCauteren K. C. Snow N. P. et al. (2019). Machine learning to classify animal species in camera trap images: Applications in ecology. Methods Ecol. Evol. 10 (4), 585–590. doi: 10.1111/2041-210X.13120
Tkachenko M. Malyuk M. Shevchenko N. Holmanyuk A. Liubimov N. (2021) Label Studio: Data labeling software [Python]. Available at: https://github.com/heartexlabs/label-studio.
Torney C. J. Lloyd-Jones D. J. Chevallier M. Moyer D. C. Maliti H. T. Mwita M. et al. (2019). A comparison of deep learning and citizen science techniques for counting wildlife in aerial survey images. Methods Ecol. Evol. 10 (6), 779–787. doi: 10.1111/2041-210X.13165
Tuia D. Kellenberger B. Beery S. Costelloe B. R. Zuffi S. Risse B. et al. (2022). Perspectives in machine learning for wildlife conservation. Nat Commun. 13 (1), 792. doi: 10.1038/s41467-022-27980-y
Wang X. Shrivastava A. Gupta A. (2017). “A-fast-rcnn: Hard positive generation via adversary for object detection,” in Proceedings of the IEEE conference on computer vision and pattern recognition. (New York City, United States of America: Institute for Electrical and Electronics Engineers (IEEE)), 2606–2615.
