Deeplearning Based Bird Deterrent System for Agriculture

Abstract

This paper addresses the solution for the problems faced by the farmers in daily life. The loss happens when the birds consume the produced crops particularl y at the time of cultivation, by which rate of production decreases. Considering those problems, the proposed system gives the solution by detecting birds automatically using convolutional neural networks. The bird detector will detect the bird when it passes through a particular range and it will compare the data with the trained images. A set of images such as bird, ball, animals and other images are trained using deep learning concept. When both the sample data and trained images are the same, it will provide loud noise. Normally birds has a sensitive hearing range, so that the bird will get disturbed As the birds have a low hearing frequency range,it gets irritated by the loud noise produced by the product and it will move from that place. By this crops can be safe and farmers can increase production rate. This works in the absence of the human.