Effective wildlife conservation requires accurate and timely monitoring of endangered species; however, conventional manual observation and analysis of camera trap imagery are labor-intensive, time-consuming, and susceptible to human error. This paper proposes a YOLOv8- based deep learning framework for automated wildlife object detection and species classification, enabling efficient identification of endangered animals from camera trap and field survey images. The proposed framework is trained and evaluated on a custom dataset comprising 1,504 highresolution images of four endangered wildlife species— Buffalo, Elephant, Rhino, and Zebra—collected from wildlife sanctuaries and conservation reserves. Images are precisely annotated using Label Studio with bounding boxes and class labels, while preprocessing techniques including image resizing, normalization, and Contrast Limited Adaptive Histogram Equalization (CLAHE) are employed to improve image quality and model robustness. A YOLOv8 Nano architecture is trained using GPU-accelerated optimization over 120 epochs to achieve high-speed and accurate detection under diverse environmental conditions. Experimental results demonstrate excellent performance, achieving an overall mAP@0.5 of 96.8%, precision of 94.3%, and recall of 92.8%, with the Rhino and Elephant classes attaining mAP@0.5 scores of 99.5% and 98.1%, respectively. Furthermore, the model performs real-time inference at approximately 4 ms per image, making it suitable for deployment in drone based surveillance, mobile applications, automated camera trap systems, and intelligent wildlife monitoring stations. The proposed framework significantly reduces manual analysis effort, improves the reliability and scalability of species monitoring, supports early detection of conservation threats such as poaching, and provides an efficient artificial intelligence-driven solution for biodiversity assessment and sustainable wildlife conservation.

Samiksha A S, Poojitha Bashiny G, Pranav D, Sabari A, Dhanya,"Wildlife Object Detection Automated Detection of Animals from Camera Trap Imagery" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(4): Page(578-583) July - August 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.