Pedestrian retrieval based on semi-coupled metric discriminative dictionary learning from sketch images

Abstract

The invention belongs to the technical field of traffic pedestrian image processing, discloses a pedestrian retrieval method and system based on semi-coupling metric discrimination dictionary learning of sketch images, establishes its own heterogeneous pedestrian retrieval database, and then performs feature extraction to extract features from images And mark different people with corresponding colors; process the extracted sample features, introduce dictionary learning method, learn dictionary pairs of heterogeneous data; learn mapping matrix from sketch image set and regular image set; introduce discriminant algorithm to learn metrics. The present invention has the advantages of solving the problem of lack of heterogeneous pedestrian data sets in the field of heterogeneous pedestrian retrieval and proposing semi-coupled metric discrimination dictionary learning (SMD) for the first time. 2 L) Technology. This technique can learn a semi-coupled mapping matrix from heterogeneous samples and dictionary pairs, which can reduce the differences between heterogeneous data to a certain extent. The ideal retrieval effect has been achieved on the new SINPID dataset.

Description

technical field [0001] The invention belongs to the technical field of traffic pedestrian image processing, in particular to a pedestrian retrieval method based on semi-coupling metric discrimination dictionary learning of sketch images. Background technique [0002] Currently, the closest prior art: [0003] Heterogeneous pedestrian retrieval between sketch images and ordinary images plays an important role in public security and criminal investigation. The purpose of heterogeneous pedestrian retrieval (HPR) is to retrieve images of the same person from heterogeneous image sets for recognition. Although pedestrian retrieval plays an important role in public safety and criminal investigation, the research is still rare. So far, in the field of pedestrian recognition, there is no pedestrian retrieval problem between sketch images and ordinary images (SINPR). data set. Therefore, it is necessary to collect a pedestrian dataset (SINPID) of sketch images and ordinary images. ...

AI Technical Summary

Problems solved by technology

The problem of pedestrian retrieval between sketch images and ordinary images that cannot be analyzed

[0008] (2) In the prior art, the semi-coupled mapping strategy is not used to connect the relationship between the sketch image and the ordinary photo to reduce the difference between heterogeneous samples

[0009] (3) No combined learning metric matrix can reveal the intrinsic projection of heterogeneous data

[0010] (4) For sketch images and ordinary photos in complex scenes, prior art dictionaries do not have good applicability to

[0011] (5) The existing technology does not combine discriminant constraints to make the same category compact and separate different categories, which is not conducive to retrieval and classification

[0020] However, directly applying existing person re-identification methods to SINPR will limit their performance

Images