Animal Detection in Farms Using OpenCV

Animal Detection in Farms Using OpenCV

Lokesh Kumar A , Aravind C , Sabareesh R , Sathiya B

ABSTRACT

A farmer cannot afford the enormous losses in agricultural earnings caused by animal incursion on their land. In the agriculture industry, computer vision is being used more and more to automate operations and increase production. Our proposal is an artificial intelligence system that uses cameras to keep an eye out for animal intrusions in the field. If an animal is spotted, the system may notify the farmer or even act independently. As the name implies, real-time object detection is the skill of identifying different items at that same moment. Identifying objects has always required courage. Faster processing power is needed to identify an item for this reason. The camera is completely utilized in this application. The camera in this application must be configured before we can identify animals. The camera will operate nonstop. The user has total control over this. The picture will be taken by the camera. The video is recorded and split up into several frames using the frame extraction modules. Using the extraction function, every frame from the movie is extracted. Usually, this extraction operates within a set time frame. If an animal is discovered in the video, the animal detection feature finds it and sounds a buzzer. But every system operating in real-time produces unlabelled data, and for effective training, a sizable collection of labeled data is needed. This work presents an application that was constructed using OpenCV libraries to detect certain things, such as animals.

Keywords: Animal detection, Farms, OpenCV, Computer vision, Real-time object detection, Python, MySQL, Intrusion detection, Buzzer alert, Frame extraction, AI-based system, Agricultural productivity.