Artificial Neural Network Based Image Processing for Wild Animal Detection and Monitoring

Abstract

Abstract-Animals leaving the forest area and entering the human habitat is increasing day by day. Animals entering the agricultural areas placed near the forest destroy crops or even attack on people therefore there is a need of system which detects the animal presence and gives warning about that in the view of security purpose. In this paper wild animal which enters the human habitation are recognized. The movement of animals in the input video is detected using Foreground detector algorithm. Extraction of foreground animal is done using Back ground subtraction based Gaussian mixture model. Morphological filters are used to remove the noise from binary image obtained from background subtraction. Training and recognition is done using back propagation algorithm. If trained images are matched with the test image, then the animal is recognized. The project aims to safe guard the wild life and animal life and it is a great help to forest department.

CNN Based Wildlife Intrusion Detection and Alert System

“…The third step is to train the CNN model using the preprocessed dataset [6]. The model architecture should include convolutional layers, pooling layers, and fully connected layers.…”