Design and Implementation of a Convolutional Neural Network-Based Anti-theft System for Reservoir Aquaculture

Abstract

This study introduces an Anti-theft system for reservoir aquaculture leveraging a convolutional neural network (CNN), comprising front-end acquisition equipment, a cloud server, and terminal monitoring devices. The system primarily employs the CNN image processing capabilities of the infrared sensor module to detect and identify unauthorized individuals accessing the reservoir. Upon detection, it issues alerts via an audio input and output module and notifies the reservoir manager through terminal monitoring equipment. The system aims to utilize CNN to efficiently provide real-time anti-intrusion updates to the reservoir manager, thereby minimizing unnecessary monitoring time and reducing labor costs.