Real Time Recording and Monitoring of Wild Animal Movements
Joel John Kandathil, Aliya Ashraf, James Ochieng Babu, Jerin Thomas, Joffin Francis · 2022 IEEE 19th India Council International Conference (INDICON) · 2022
Animal attacks are one of the major threats that we face today. As the human population and resources need to grow, we humans are leaving less and less room for wildlife, which has increased animal attacks in many human-populated places worldwide. Animals are losing their homes due to globalisation and industrialisation, and they are forced to seek sanctuary and food in human-populated areas. Wildlife encroachment is harmful to humans and animals in areas with a high population and high human mobility. Through this paper, we intend to safeguard the life of wild animals as well as humans. We try to answer the problem by combining deep learning approaches from many computer vision domains, such as object detection. Here we use the YOLOv5 object detection model to ascertain the presence of wild animals in images. Once detected, the animals’ intentions are found, and different frequency of sound, depending on the species of the animal, is used to terrify them back to their habitats; simultaneously, alerts are sent to the residents of the village.