Research on Detection of Floating Objects on Water Surface Based on Faster-RCNN

With the development of industry, water environmental pollution is becoming increasingly severe, leading to a decrease in water quality and posing a threat to human production, life, and economic and social development. Floating debris on the water surface, as one of the important reasons for polluting water resources, has become an urgent environmental safety issue to be solved. The traditional manual river patrol method has difficulties and inefficiencies in detecting and handling floating objects on the water surface, so it is necessary to study an intelligent detection and recognition method for floating objects on the water surface. This article constructs a dataset of floating objects on the water surface in the river patrol scene, and conducts research on floating object detection based on Faster-RCNN method. Convolutional neural network models such as VGG16-RCNN, MobileNetV2-RCNN, and ResNet50-RCNN are used. The experimental results show that the Faster-RCNN method has high accuracy and efficiency in detecting floating objects on the constructed data, and can effectively address water pollution issues.