Neural Network Based Image Classification for Animal Intrusion Detection System

Abstract

Every year a large number of resources are lost and human life gets in danger due to human-animal clashing in the forest zone and agriculture fields.  People end up losing their livestock, crop, property and sometimes their lives too. To prevent the entry of wild animals in the farm field, it has to be monitored all the time. Here we come with a system which will monitor the field for us. For the first round of surveillance, we have a PIR sensor for motion detection and a video camera to take pictures of the animals. In the next stage, we send these pictures of the animals to the Raspberry Pi for processing. These pictures are processed and classified using deep neural network. Even animal counting is done. Finally, the picture is analyzed to check intrusion of wild animal so that appropriate action can be taken based on the type of animal.