Target re-recognition method and device and computer readable storage medium

Abstract

The invention discloses a target re-recognition method and device and a computer readable storage medium. The method comprises the steps: obtaining a target image and a to-be-retrieved video, and thetarget image comprises a retrieval target; performing target detection on the to-be-retrieved video to obtain an illumination intensity value corresponding to the to-be-retrieved video; determining asimilarity threshold matched with the illumination intensity value by utilizing the illumination intensity value; calculating the similarity between the retrieval target and each target in the to-be-retrieved video, and judging whether the similarity is greater than a similarity threshold value or not; and if the similarity between the retrieval target and the target in the to-be-retrieved video is greater than a similarity threshold, determining that the retrieval target is in the to-be-retrieved video. In this way, the accuracy of target recognition can be improved.

Description

technical field [0001] The present application relates to the technical field of image processing, and in particular to a target re-identification method, device and computer-readable storage medium. Background technique [0002] In the actual monitoring scene, due to the complex environment and low resolution of the camera, it is difficult to obtain a clear face, and the role of face recognition technology is not prominent; as a supplement to face recognition, pedestrian re-identification can effectively solve the problem in practical applications. Cross-device or cross-scenario issues. Pedestrian re-identification needs to identify whether the pedestrians in different environments and lighting conditions are the same person. The possible recognition scenes include different lighting environments such as daytime, night, cloudy, weak light or strong light. The recognition effect is greatly affected, and the current pedestrian re-identification methods do not take into accoun...

AI Technical Summary

Problems solved by technology

[0002] In the actual monitoring scene, due to the complex environment and low resolution of the camera, it is difficult to obtain a clear face, and the role of face recognition technology is not prominent; as a supplement to face recognition, pedestrian re-identification can effectively solve the problem in practical applications. Cross-device or cross-scenario issues. Pedestrian re-identification needs to identify whether the pedestrians in different environments and lighting conditions are the same person. The possible recognition scenes include different lighting environments such as daytime, night, cloudy, weak light or strong light. The recognition effect is greatly affected, and the current pedestrian re-identification methods do not take into account that the retrieved images may come from different environments, and the retrieval effect is reduced due to different lighting conditions. Therefore, a re-identification method that solves the environmental impact is urgently needed

Images