Video image feature extraction method for police drones investigation and evidence collection

In the practical application of police drones in investigation and evidence collection, it is not only necessary to operate stably under unfavorable conditions such as night, haze, and complex terrain, but also to ensure the operation continuity and reliability. The drones needs to have good compati...

Full description

Saved in:
Bibliographic Details
Main Authors: CHEN Sixin, HE Fengwei
Format: Article
Language:zho
Published: Editorial Office of Journal of XPU 2024-12-01
Series:Xi'an Gongcheng Daxue xuebao
Subjects:
Online Access:http://journal.xpu.edu.cn/en/#/digest?ArticleID=1523
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849321867578441728
author CHEN Sixin
HE Fengwei
author_facet CHEN Sixin
HE Fengwei
author_sort CHEN Sixin
collection DOAJ
description In the practical application of police drones in investigation and evidence collection, it is not only necessary to operate stably under unfavorable conditions such as night, haze, and complex terrain, but also to ensure the operation continuity and reliability. The drones needs to have good compatibility and be able to seamlessly integrate with other police systems to achieve information sharing and collaborative operations. In response to these requirements, a new video image feature extraction method was proposed in the article. Firstly, we made a detailed analysis of the HSV color of the target object in the video image. Then, based on the results of the HSV color space, attention mechanism was introduced for video image feature extraction, and constructed spectral multi-scale feature extraction network and spatial multi-scale feature extraction network with attention mechanism. Finally, the spectrum and spatial network were merged into a whole, and trained and fused to obtain a spectral spatial joint feature extraction network, which achieved the extraction for police unmanned aerial vehicle investigation and evidence collection. The experimental results show that the feature deviation of the method in the article is relatively small, and the highest feature extraction accuracy is 97%. The overall frame per second (FPS) value is between 50~75, which remains at a high level and has good stability and reliability. This not only meets the demand for efficient and accurate feature extraction of police drones, but also provides technical support for their widespread application in complex investigation environments.
format Article
id doaj-art-64d2b3a12c534301858eb4b4de43f3f6
institution Kabale University
issn 1674-649X
language zho
publishDate 2024-12-01
publisher Editorial Office of Journal of XPU
record_format Article
series Xi'an Gongcheng Daxue xuebao
spelling doaj-art-64d2b3a12c534301858eb4b4de43f3f62025-08-20T03:49:37ZzhoEditorial Office of Journal of XPUXi'an Gongcheng Daxue xuebao1674-649X2024-12-013869810410.13338/j.issn.1674-649x.2024.06.013Video image feature extraction method for police drones investigation and evidence collectionCHEN Sixin0HE Fengwei1Police Tactics Department, Chongqing Police College, Chongqing 401331, ChinaPolice Tactics Department, Chongqing Police College, Chongqing 401331, ChinaIn the practical application of police drones in investigation and evidence collection, it is not only necessary to operate stably under unfavorable conditions such as night, haze, and complex terrain, but also to ensure the operation continuity and reliability. The drones needs to have good compatibility and be able to seamlessly integrate with other police systems to achieve information sharing and collaborative operations. In response to these requirements, a new video image feature extraction method was proposed in the article. Firstly, we made a detailed analysis of the HSV color of the target object in the video image. Then, based on the results of the HSV color space, attention mechanism was introduced for video image feature extraction, and constructed spectral multi-scale feature extraction network and spatial multi-scale feature extraction network with attention mechanism. Finally, the spectrum and spatial network were merged into a whole, and trained and fused to obtain a spectral spatial joint feature extraction network, which achieved the extraction for police unmanned aerial vehicle investigation and evidence collection. The experimental results show that the feature deviation of the method in the article is relatively small, and the highest feature extraction accuracy is 97%. The overall frame per second (FPS) value is between 50~75, which remains at a high level and has good stability and reliability. This not only meets the demand for efficient and accurate feature extraction of police drones, but also provides technical support for their widespread application in complex investigation environments.http://journal.xpu.edu.cn/en/#/digest?ArticleID=1523police dronesvideo image feature extractionattention mechanismhsv color space
spellingShingle CHEN Sixin
HE Fengwei
Video image feature extraction method for police drones investigation and evidence collection
Xi'an Gongcheng Daxue xuebao
police drones
video image feature extraction
attention mechanism
hsv color space
title Video image feature extraction method for police drones investigation and evidence collection
title_full Video image feature extraction method for police drones investigation and evidence collection
title_fullStr Video image feature extraction method for police drones investigation and evidence collection
title_full_unstemmed Video image feature extraction method for police drones investigation and evidence collection
title_short Video image feature extraction method for police drones investigation and evidence collection
title_sort video image feature extraction method for police drones investigation and evidence collection
topic police drones
video image feature extraction
attention mechanism
hsv color space
url http://journal.xpu.edu.cn/en/#/digest?ArticleID=1523
work_keys_str_mv AT chensixin videoimagefeatureextractionmethodforpolicedronesinvestigationandevidencecollection
AT hefengwei videoimagefeatureextractionmethodforpolicedronesinvestigationandevidencecollection