Early Warning System from Threat of Wild Animals using Digital Image Processing

Abstract

Wildlife infiltration in places with high human mobility has been proven to be dangerous for both humans and animals. If people fail to recognize an approaching wild animal, it may result in a direct attack. Due to their size and style of movement, monitoring and surveillance of wild animals is challenging. Additionally, it is a significant task to identify the species that were photographed. Elephants, tigers, and monkeys pose a serious threat to humans, and it will take a very long time for them to recover. Because interactions between humans and animals can be harmful to both species, successive frame differencing makes it possible to identify moving objects in videos. By utilizing the traits, the moving objects can be identified. Interactions between humans and animals can be hazardous. The proposed system is based on digital image processing, convolutional neural network and background subtraction method.