Human body cover generation method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a human body cover generation method and device, electronic equipment and a storage medium, and the method comprises the steps: filtering out human body images with low quality according to a quality score, and filtering out human body images with low quality, such as low definition, severe shielding or truncation, dark illumination, exposure and the like. The filtered human body images are clustered based on the similarity matrix and the human body features, and the human body image of the specified personnel type in the clustering result is output as the human body cover, so that the human body image with the wrong classification, the human face region image and the human body image with the wrong association of the human face region image can be effectively filtered out; therefore, errors of the human body cover are reduced, and the accuracy of the generated human body cover is improved. Moreover, the method can be used in various scenes containing personnel files, can adapt to different purification requirements through the preset score threshold and the selection of specified personnel types, and is good in applicability.

Description

technical field [0001] The present application relates to the technical field of image processing, in particular to a human body cover generation method, device, electronic equipment and storage medium. Background technique [0002] With the improvement of people's safety awareness and the popularization of surveillance cameras, artificial intelligence surveillance technology based on computer vision has developed rapidly, and the establishment of personnel files has gradually become a key deployment direction, which is of great significance for maintaining personnel and property safety. The personnel files have also changed from traditional face-based files to combined face-body files. [0003] In the related art, after the face area image and the human body area image are detected, the face area image and the human body area image are directly associated, and a file is created to obtain a personnel file. When the personnel file needs to be used, output the cover of the hu...

AI Technical Summary

Problems solved by technology

However, for scenes where there are multiple people in the same image or occluded objects in the image, it will cause errors in the association between the image of the face area and the image of the body area, resulting in an error in the generated body cover. In addition, due to the accuracy of the face recognition algorithm, It will cause the image error of the face area in the personnel file (that is, the face of person A is divided into the personnel file of person B), and it will also cause the error of the generated human body cover, which will affect the accuracy of subsequent behavior prediction and other operations.

Images