Bird repeller for agriculture lands and orchards.

The economy of the countries like India, China, Nepal, and Indonesia has a huge dependency on agriculture. So it is important to increase or keep a stable production of food commodities like rice, wheat, corn, barley, and fruits like apples, berries, grapes in vineyards and etc. But apart from climatic changes and plant disease bird pest is one of the problems in crop field and fruit farms. Because every year a considerable amount of production of crops and fruits is consumed and wasted by birds. The existing systems are utilizing motion sensors, IR sensors, and PIR sensors are used in the existing systems. The problem with the existing system they can't identify the object. Because of that no matter what interferes with the sensors it will alarm and starts sounding. So this paper has been providing a solution to the problems of existing systems in the way of neural networks and deep learning to protect the grains and fruits in the farming environment from birds. In this method, the system utilizes a visual unit, power unit, acoustic speaker unit, and control and processing unit. The processing unit contains a program that is based on the DNN object detection algorithm and a dataset called COCO-Common Objects in Context. With the help of peripherals and programs, it can able to detects birds in its range. The DNN Detection model is detecting the moving subject. If any birds are detected from the video feed a threat sound is broadcasted throughout the surrounding. This could be an effective solution for bird pest problems in agriculture fields. [ABSTRACT FROM AUTHOR]