Animal Motion Tracking in Forest: Using Machine Vision Technology
Bhaskar Borah1, Ria Saikia2, Piyali Das3
1B. Tech Final Year Student, 2-3B.Sc Final Year Student
1Department of Electronics and Telecommunication Engineering, Assam Engineering College, India
2-3Department of Information Technology, The Assam Royal Global University, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The safety of wildlife is affected by various poaching done in the forests. Rhino poaching in Assam is one of the major environmental issues in India which continues in the region of Kaziranga National Park, Manas National Park, and some other grasslands of Assam. Indian rhinoceros inhabited most of the floodplain of the Indogangetic and Brahmaputra riverine tracts and the neighboring foothills. This also includes poaching of other animals like deer, tiger, etc. Despite taking preventive measures to curb the poaching of rhinos and protect all animals in the Unesco world heritage site Kaziranga National Park and Tiger Reserve (KNP and TR), the results have not shown much impact on the situation and manual observation of animal motion activity is time and cost intensive. Therefore automatic detection and monitoring of live captive animals is of major importance for assessing animal activity and, thereby, allowing for early recognition of changes indicative for threats, diseases, and animal welfare issues. To the end, this project gives a much more accurate technique to track animal motions and ensure the safety of animals. We use computer vision, being a non-invasive method for the automatic monitoring of animals. More specifically, we are using YOLO v4 (You Only Look Once version 4) model for detecting animals along with Deep SORT ( Simple Online and Realtime Tracking with a Deep Association Metric ) for animal motion tracking. Computer vision thereby outperforms manual and sensor-based exhaustive monitoring of the animals. This in turn can also be used for animal behavioral analysis and thus for real-time animal monitoring.
Key Words: Motion tracking, Animal detection, YOLO, Deep SORT
