Improving Person Re-Identification Rate in Security Cameras by Orthogonal Moments and a Distance-based Criterion
Surveillance and security cameras help security forces in public places such as airports, railway stations, universities and office buildings to perform high-level surveillance tasks such as detecting suspicious activity or anticipating undesirable events. Re-Identification (Re-ID) is defined as the...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
OICC Press
2024-02-01
|
| Series: | Majlesi Journal of Electrical Engineering |
| Subjects: | |
| Online Access: | https://oiccpress.com/mjee/article/view/4898 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850283627578916864 |
|---|---|
| author | Ali Dadkhah Saeed Nasri |
| author_facet | Ali Dadkhah Saeed Nasri |
| author_sort | Ali Dadkhah |
| collection | DOAJ |
| description | Surveillance and security cameras help security forces in public places such as airports, railway stations, universities and office buildings to perform high-level surveillance tasks such as detecting suspicious activity or anticipating undesirable events. Re-Identification (Re-ID) is defined as the process of communicating between images of the person in different cameras in a surveillance environment. Changing the field of view of any camera presents challenges such as changing body posture, changing brightness, noise and blockage. This article focuses on extracting the most distinctive features to overcome these challenges. The features of Hu moment, Zernike moment in 9th order and Legendre moment in 9th order for each image are extracted and merged into a single feature vector to form a single feature vector for each image. Principal Component Analysis (PCA) was used to reduce the vector dimensionality and finally the Mahalanobis distance criterion was used for identification. The proposed method in the VIPeR database has achieved a re-ID rate of 96.5. Although the presented method is simple, the outcome has been superior compared to many of the state-of-the-art methods. |
| format | Article |
| id | doaj-art-3aad0b2269144d87a0d4e4beebc63e7d |
| institution | OA Journals |
| issn | 2345-377X 2345-3796 |
| language | English |
| publishDate | 2024-02-01 |
| publisher | OICC Press |
| record_format | Article |
| series | Majlesi Journal of Electrical Engineering |
| spelling | doaj-art-3aad0b2269144d87a0d4e4beebc63e7d2025-08-20T01:47:44ZengOICC PressMajlesi Journal of Electrical Engineering2345-377X2345-37962024-02-0114410.29252/mjee.14.4.85Improving Person Re-Identification Rate in Security Cameras by Orthogonal Moments and a Distance-based CriterionAli Dadkhah0Saeed Nasri1Department of Computer Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran.Department of Computer Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran.Surveillance and security cameras help security forces in public places such as airports, railway stations, universities and office buildings to perform high-level surveillance tasks such as detecting suspicious activity or anticipating undesirable events. Re-Identification (Re-ID) is defined as the process of communicating between images of the person in different cameras in a surveillance environment. Changing the field of view of any camera presents challenges such as changing body posture, changing brightness, noise and blockage. This article focuses on extracting the most distinctive features to overcome these challenges. The features of Hu moment, Zernike moment in 9th order and Legendre moment in 9th order for each image are extracted and merged into a single feature vector to form a single feature vector for each image. Principal Component Analysis (PCA) was used to reduce the vector dimensionality and finally the Mahalanobis distance criterion was used for identification. The proposed method in the VIPeR database has achieved a re-ID rate of 96.5. Although the presented method is simple, the outcome has been superior compared to many of the state-of-the-art methods.https://oiccpress.com/mjee/article/view/4898Genetic AlgorithmMahalanobis distanceOrthogonal MomentsPerson re-identification |
| spellingShingle | Ali Dadkhah Saeed Nasri Improving Person Re-Identification Rate in Security Cameras by Orthogonal Moments and a Distance-based Criterion Majlesi Journal of Electrical Engineering Genetic Algorithm Mahalanobis distance Orthogonal Moments Person re-identification |
| title | Improving Person Re-Identification Rate in Security Cameras by Orthogonal Moments and a Distance-based Criterion |
| title_full | Improving Person Re-Identification Rate in Security Cameras by Orthogonal Moments and a Distance-based Criterion |
| title_fullStr | Improving Person Re-Identification Rate in Security Cameras by Orthogonal Moments and a Distance-based Criterion |
| title_full_unstemmed | Improving Person Re-Identification Rate in Security Cameras by Orthogonal Moments and a Distance-based Criterion |
| title_short | Improving Person Re-Identification Rate in Security Cameras by Orthogonal Moments and a Distance-based Criterion |
| title_sort | improving person re identification rate in security cameras by orthogonal moments and a distance based criterion |
| topic | Genetic Algorithm Mahalanobis distance Orthogonal Moments Person re-identification |
| url | https://oiccpress.com/mjee/article/view/4898 |
| work_keys_str_mv | AT alidadkhah improvingpersonreidentificationrateinsecuritycamerasbyorthogonalmomentsandadistancebasedcriterion AT saeednasri improvingpersonreidentificationrateinsecuritycamerasbyorthogonalmomentsandadistancebasedcriterion |