Pedestrian re-identification method and a related product

Abstract

The invention provides a pedestrian re-identification method and a related product. The method comprises the steps of conducting feature extraction on a target image through a preset convolutional neural network training model, acquiring a first feature set,wherein the preset convolutional neural network training model is composed of a first training module and a second training module, and features extracted by the first training module and the second training module are fused into a feature set; Determining a Hamming distance between the first feature set and each second feature set in the plurality of second feature sets to obtain a plurality of Hamming distance values; Calculating a similarity probability value between the input image and each image in the image library through the plurality of Hamming distance values to obtain a plurality of similarity probability values; Selecting a similarity probability value greater than a preset threshold value from the plurality of similarity probability values to obtain at least one target similarity probability value; And displaying the images in the image library corresponding to the at least one target similarity probability value tothe user according to the descending order of the similarity probability values. The pedestrian recognition method and device can improve the pedestrian recognition accuracy.

Description

technical field [0001] This application relates to the technical field of image processing, in particular to a pedestrian re-identification method and related products. Background technique [0002] Person Re-identification (Person Re-identification), referred to as ReID, is a technology that uses computer vision technology to determine whether a specific pedestrian exists in an image or video sequence, that is, given a monitored pedestrian image, retrieve the cross-device target image. It is designed to compensate for the visual limitations of current fixed cameras, and can be combined with pedestrian detection / pedestrian tracking technology. [0003] Pedestrian re-identification (ReID), considered as a sub-problem of image retrieval, can be widely used in intelligent video surveillance, intelligent security and other fields. In surveillance video, due to the resolution of the camera, it is usually impossible to obtain very high-quality face pictures. In the case of face...

AI Technical Summary

Problems solved by technology

Moreover, the actual video surveillance scene is complex, and the angles of different cameras, lighting and other environmental influences lead to the appearance characteristics of the same pedestrian under different cameras will be different; on the contrary, due to changes in camera angles and pedestrian postures , leading to the appearance characteristics of different pedestrians may be more similar than the appearance characteristics of the same person, which in turn has a great impact on pedestrian re-identification

Images