Real-time detection method of surface floating objects based on deep learning

Abstract

In order to solve the problem that the surface floating object needs manual inspection and time-consuming and labor-intensive problems, and the information of single monitoring means is not comprehensive, this paper proposes a set of integrated monitoring and detection system, which can monitor the video image information in various scenarios. Automatically alert and dispose of. Based on the video surveillance-based surface floating object detection algorithm, the darknet framework is used to establish a deep learning network, and the improved YOLOv3 detection algorithm is designed to solve the problem that the garbage floating on the fast flowing water surface and the algae and other pollutants cannot be quickly identified.