The Residential Rat Monitoring and Detection System based on Object Detection

With the popularity of automatic monitoring systems in home environments, animal detection techniques based on computer vision have gradually become a research hotspot. In this article, a rat detection method based on the improved YOLOv5 algorithm is proposed, aiming to improve the detection accuracy of the model in complex environments, especially in low-light and occlusion situations. Firstly, to address the limitations of YOLOv5 in small object detection, this article employs data enhancement techniques and incorporates high-resolution image inputs to improve the detection of rat, a small target. Second, a post-processing method based on adaptive thresholding is proposed to reduce the false detection of low-confidence targets while enhancing the robustness of the model in complex backgrounds. In addition, the feature extraction capability of YOLOv5 is optimised by introducing a new attention mechanism model, which improves the accuracy of rats detection. The experimental results show that the improved YOLOv5 model exhibits higher accuracy and stronger robustness in the rats detection task, providing an effective solution for mouse monitoring in home environments.