Pedestrian re-identification method based on multi-characteristic integration

Abstract

The invention discloses a pedestrian re-identification method based on multi-characteristic integration; the method comprises the following steps: 1, collecting multi-characteristic information; 2, carrying out tolerance model learning on the multi-characteristic information collected by step1, and building a plurality of coupling models based on a single-characteristic; 3, weight learning: distributing a corresponding weight for each model according to the identification level presentation of the 4 models learned in step2 on a training data set; 4, multi-characteristic distribution: using each single-characteristic coupling model obtained in step2 to calculate an assigned image and a coupling image, thus obtaining the distance between images on each single-characteristic coupling model, ranking the distances to obtain a corresponding coupling result, carrying out weight integration for distances on each characteristic and weights obtained by step3 so as to solve the final combined distance, and further obtaining an optimized multi-characteristic coupling result.

Description

technical field [0001] The invention belongs to the technical field of image retrieval and relates to a pedestrian re-identification method based on multi-feature integration. Background technique [0002] Pedestrian re-identification refers to identifying a specific pedestrian who has appeared in the image library in the possible image library. For example, a typical application scenario of pedestrian re-identification technology is criminal investigation. Given a suspect's photo, the video surveillance video related to the suspect is collected, and then clues related to the suspect are collected. With people's increasing attention to social public safety and the development of video capture technology and large-scale data storage technology, a large number of surveillance cameras are used in shopping malls, parks, schools, hospitals, subway stations and other places with dense crowds and prone to public safety incidents. , The emergence of surveillance cameras has undoubt...

AI Technical Summary

Problems solved by technology

[0012] In order to achieve the above purpose, the present invention provides a pedestrian re-identification method based on multi-feature integration, which solves the problem in the prior art that the traditional recognition algorithm cannot obtain relatively accurate data due to the complex dynamic characteristics and instability of the data. The problem of ideal results improves the accuracy of pedestrian re-identification

Images