Item illegal throwing detection method and device, server and storage medium

Abstract

The embodiment of the invention discloses an item illegal throwing detection method and device, a server and a storage medium. The item illegal throwing detection method comprises the steps of acquiring an item sorting monitoring video; performing semantic segmentation processing on multiple frames of monitoring images in the monitoring video to obtain a throwing image; determining the throwing direction of the item from the throwing image; determining a throwing area of the item in the throwing image according to the throwing direction; and detecting the throwing type of the item in the throwing area to determine whether the item is illegally thrown. According to the item illegal throwing detection method, semantic segmentation processing is carried out on the frame monitoring image in the monitoring video, the throwing image containing the item and the person can be accurately obtained, then the throwing direction and the throwing area of the item are determined according to the throwing image, the throwing track of the item in the throwing area is detected, whether the item is thrown illegally or not can be detected more accurately, and the accuracy of the method for detecting the illegal throwing of the item is improved.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a method, device, server and storage medium for detecting illegal throwing of articles. Background technique [0002] In recent years, with the development of the logistics industry, in order to pursue timeliness, sorting personnel may have violent sorting behaviors in the process of sorting items, such as throwing or kicking items. These behaviors are easy to Items cause damage, resulting in huge compensation for the courier company. For this reason, the existing technology proposes a method of using surveillance video detection technology to perform regression classification on the target area in the surveillance video through target detection technology, so as to judge whether the item sorting personnel in the surveillance video have the behavior of throwing items in violation of regulations, but , the accuracy of this detection technology is low and cannot meet busin...

AI Technical Summary

Problems solved by technology

[0002] In recent years, with the development of the logistics industry, in order to pursue timeliness, sorting personnel may have violent sorting behaviors in the process of sorting items, such as throwing or kicking items. These behaviors are easy to Items cause damage, resulting in huge compensation for the courier company

For this reason, the existing technology proposes a method of using surveillance video detection technology to perform regression classification on the target area in the surveillance video through target detection technology, so as to judge whether the item sorting personnel in the surveillance video have the behavior of throwing items in violation of regulations, but , the accuracy of this detection technology is low and cannot meet business needs

Images