Wildlife Detection: A Cost Effective Approach

Wildlife detection requires human effort which can be minimised with the help of deep learning which is a subset of artificial intelligence. Wildlife detection is useful for wide range of applications such as protection of crops, population density of wildlife, prevention of occupation of wildlife areas. This project helps in realising mentioned applications with the help of YOLO (You Only Look Once) object detection framework. Dataset consisting of wildlife images is used for training and trained model achieved 86.32% precision, 75.09% recall and F1 score is 80.31. After training, model is deployed in Raspberry Pi 4 model B. Raspberry Pi is lightweight, low consumption of power, portable and hence it could be used for detecting the presence of wildlife in real time. YOLOv8 is one of the latest versions of YOLO which has better accuracy and speed compared to previous versions of YOLO and uses less floating-point operations per second.