A Garbage Recognition Algorithm Based on YoloX Using in Flood Control System

In recent years, floods in city have occurred frequently, causing great losses to the personal safety and property safety of city dwellers. One of the main factors causing flood disasters in city is that the garbage is not cleaned up in time, resulting in the blockage of the sewer manhole cover and the failure to discharge sewage in time. The daily inspection of urban environment and the garbage clean-up work are often powerless in the face of enormous pressure in sanitation work. Compared with relying on manual screening, real-time monitoring of manhole covers through road cameras, and data will be fed back timely once the garbage accumulation is identified, which can effectively improve the efficiency of urban sanitation work. This paper uses the data captured by urban surveillance cameras, identifies manhole covers and garbage using the algorithm based on the YoloX. And this paper builds a flood preventing system that can automatically monitor and identify the accumulation of garbage and send the data to the client for data display and giving early warning.