A pedestrian rerecognition method based on multi-view image feature decomposition

Abstract

The invention relates to the technical field of intelligent image retrieval, More specifically, the invention relates to a pedestrian recognition method based on multi-view image feature decomposition, which starts from the problem of target image classification and combines with capsule network to construct a pedestrian image multi-view image feature generation network, and decomposes any pedestrian image to obtain multi-view image features and similarity under the view. Features of the same view angle are directly used for re-recognition, and features of different view angles must undergo feature conversion. In the invention, a feature conversion matrix obtained by using a BP neural network solves the measurement problem of features of different view angles. Experiments show that the decomposition of pedestrian images is very helpful to improve the accuracy of pedestrian recognition.

Description

technical field [0001] The invention relates to the technical field of intelligent image retrieval, and more particularly, to a pedestrian re-identification method based on multi-view image feature decomposition. Background technique [0002] With more and more attention paid to social public safety, a large number of surveillance cameras have been installed in many important places. The surveillance videos captured by surveillance cameras can provide the public security departments with clues to detect major criminal cases such as terrorist attacks and crowd fights. Pedestrian re-identification is an automatic target recognition technology, which can quickly locate the human target of interest in the surveillance network, and is an important step in applications such as intelligent video surveillance and human behavior analysis. The main purpose of using pedestrian re-identification technology is to solve the problem of pedestrian retrieval in the field of security monitori...

AI Technical Summary

Problems solved by technology

From the research results described above, it can be seen that the research on image-based pedestrian re-identification has made significant progress, but there is room for further improvement, mainly due to the problem of low accuracy.

Images