Paper Title
ANIMAL FOOTPRINT CLASSIFICATION WITH CNNs

Abstract
The steady decline in global biodiversity highlights an urgent need for smarter and more efficient methods to monitor and protect wildlife. Conventional techniques like camera traps and manual footprint tracking, while useful, are often slow, labor-intensive, and difficult to scale across large natural habitats. To overcome these limitations, this work introduces an automated system that uses computer vision and deep learning to accurately detect and classify animal footprints. The system is powered by Convolutional Neural Networks (CNNs) trained on a wide and diverse collection of footprint images. This enables it to identify animal species with high reliability while greatly reducing the need for human intervention. Experimental evaluations show that the proposed model achieves an accuracy of over 90 By offering a non-invasive, affordable, and scalable approach, the system can significantly enhance wildlife research and conservation activities. It supports ecologists and forest officers in monitoring animal populations, understanding habitat use, and preventing illegal poaching. In the future, the dataset can be expanded to include rare species, and real-time, edge-based processing can be integrated to support on-site wildlife monitoring in remote environments. Keywords - Wildlife Monitoring, Biodiversity Conservation, Deep Learning, Convolutional Neural Networks (CNN), Animal Footprint Detection, Computer Vision, Edge AI.