Pedestrian re-identification method based on multi-region feature extraction and fusion

Abstract

The invention relates to a pedestrian re-identification method based on multi-region feature extraction and fusion, including the following steps: using a residual network to extract global features,and adding a pedestrian identity classification module for global feature extraction and optimization in the training stage; constructing a multi-region feature extraction sub-network for local feature extraction, and carrying out weighted fusion of local features; setting a loss function including the loss of the classification module and the loss of a feature fusion module; training the networkto obtain a model, and extracting the feature vectors of a query set and a test set; and in the measuring stage, re-measuring the feature distance by using a cross-nearest neighbor method. The methodis reasonable in design. Global and local feature are effectively combined. The distance measurement method is optimized. A good pedestrian re-identification result is obtained. The overall matching accuracy of the system is greatly improved.

Description

technical field [0001] The invention belongs to the technical field of computer vision pedestrian re-identification, in particular to a pedestrian re-identification method based on multi-region feature extraction and fusion. Background technique [0002] With the continuous improvement of video capture technology and the rapid development of large-scale data storage technology, it is possible to apply a large number of surveillance camera systems in public places. The identification and processing of pedestrians in massive surveillance videos presents a trend of rapid development. Relying only on human eyes to identify pedestrians in surveillance images is obviously very inefficient. The task of pedestrian re-identification technology is to rely on computer vision technology to solve non-overlapping surveillance fields of view. The problem of identity matching of pedestrians. [0003] There are two main categories of traditional solutions: one is feature extraction, which i...

AI Technical Summary

Problems solved by technology

[0004] However, most of the existing convolutional neural networks extract the features of the whole picture, and do not make full use of the local features of the picture. These local features are more robust to the pose changes of pedestrians under different viewing angles, and are more effective for distinguishing different pedestrians. , therefore, the overall matching accuracy of existing person re-identification methods is not high

Images