Residual Waste Quality Detection Method Based on Gaussian-YOLOv3

In the process of garbage collection, the water flow in residual waste directly affects the process of collecting residual waste. Therefore, detecting the water flow in residual waste at the garbage transfer station is of great guiding significance for garbage disposal. In this paper, the Gaussian-YOLOv3 algorithm with high accuracy and real-time performance is used to identify and detect the water flow during the dumping process of residual waste, and determine the quality of the classification of residual waste according to the recognition situation. The experimental results show that the residual waste quality detection method based on the Gaussian-YOLOv3 algorithm can accurately identify the amount of the water flow during the dumping of the residual waste. At the same time, the back annotation and retraining method significantly reduces the model's impact on similar residual waste in complex environments. The false recognition rate satisfies the actual needs of residual waste water flow identification and improves the efficiency of residual waste classification quality determination.