Wild Animals Detection and Alert System Using Deep Learning
Dandao Xuebao/Journal of Ballistics · 11.06.2026 · 38 (2) · 372-383
DOI: 10.52783/dxjb.v38.353 ↗Kurz gefasst
Fachartikel in der Zeitschrift Dandao Xuebao/Journal of Ballistics, online erschienen am 11.06.2026. Verfasst von Mayur Asaram Dahare.
Abstract (Originalsprache)
Human-wildlife conflicts have become one of the greatest conservation and safety issues in Asia and Africa and have resulted in thousands of deaths and millions of dollars in agricultural damages each year. In this paper, we propose WADAS (Wild Animal Detection and Alert System) which is a real-time deep learning-based approach to detect wild animals in proximity to human settlements and provide instant alerts through multiple channels to reduce the risk of conflict. We have implemented a fine-tuned YOLOv8 object detection backbone, a custom convolutional neural network (CNN) classifier and deployed them onto edge computing hardware, which is connected to infrared camera traps and PIR sensors. It is able to reach the mean average precision (mAP@0.5) of 94.7% on 18 wild animal species such as turtles, snakes, wild boars, elephant and leopard. The real-time version of the system runs on NVIDIA Jetson Nano and achieves 28.4 frames per second, and an average alert latency of 1.2 seconds. The system was deployed in the field over 14 months in six villages in the forest-fringe areas of Uttarakhand, India and it has been effective in verifying 4,312 detections and warning local villages of the high risk in 97.3% of the encounters. By comparing the results, it can be seen that WADAS is more precise (94.7% vs. 87.3%), more recall (92.1% vs. 83.6%) than existing systems and has a faster response time. This work provides a carefully curated collection of 47,500 annotated wildlife images from various illumination and weather conditions that is made open-sourced for use by the research community.
Zitieren
APA
Dahare, M. A. (2026). Wild Animals Detection and Alert System Using Deep Learning. Dandao Xuebao/Journal of Ballistics, 38(2), 372-383. https://doi.org/10.52783/dxjb.v38.353
Vancouver
Dahare MA. Wild Animals Detection and Alert System Using Deep Learning. Dandao Xuebao/Journal of Ballistics. 2026;38(2):372-383. doi:10.52783/dxjb.v38.353
BibTeX
@article{Dahare2026,
author = {Dahare, Mayur Asaram},
title = {Wild Animals Detection and Alert System Using Deep Learning},
journal = {Dandao Xuebao/Journal of Ballistics},
year = {2026},
volume = {38},
number = {2},
pages = {372--383},
doi = {10.52783/dxjb.v38.353}
}
EUVIMED