Poultry Farming Abnormality Detection using ESP32-CAM and OpenCV

Abstract

This research studies the use of the ESP32-CAM module when combined with SSD MobileNet and OpenCV to detect abnormalities in chicken farm operations, including feather plucking and pecking caused by imprisonment. The ESP32-CAM works as a video streaming server, providing real-time monitoring and analysis. Two techniques are investigated: the first includes the ESP32-CAM acting as a video streaming server for browser clients, while OpenCV and SSD MobileNet process the video stream and send the results back to the server. A prototype user interface is constructed for management. The second option uses the ESP32-CAM as a video streaming server for Python clients, with the video stream being processed on the client side using SSD MobileNet and OpenCV.  This dual strategy guarantees that poultry farm settings are securely monitored and controlled, while also increasing management efficiency and resolving behavioral concerns.