基于三維空間的多行人重識別方法

打開文本圖片集

摘要：為了解決監(jiān)控中的遮擋問題和提高人群監(jiān)控的準(zhǔn)確性，該文提出了基于三維空間的多行人重識別方法，結(jié)果表明，該方面可以大幅提高監(jiān)控系統(tǒng)的覆蓋范圍和人群監(jiān)控系統(tǒng)在復(fù)雜環(huán)境下的識別與跟蹤準(zhǔn)確性，有效支持密集人群環(huán)境下的安全監(jiān)控需求。

關(guān)鍵詞：三維空間；行人重識別；人群監(jiān)控

doi：10.3969/J.ISSN.1672-7274.2024.07.021

中圖分類號：TP 3                           文獻標(biāo)志碼：A            文章編碼：1672-7274（2024）07-00-03

Multi-pedestrian Rerecognition Method Based on Three-dimensional Space

SUN Yiyang

（Hubei University. Wuhan 430062， China）

Abstract： In order to solve the occlusion problem and improve the accuracy of crowd monitoring， a multi-pedestrian rerecognition method based on three-dimensional space is proposed. The research results show that the coverage of the monitoring system and the identification and tracking accuracy of the crowd monitoring system in complex environments can be greatly improved， and the security monitoring requirements in dense crowd environments can be effectively supported.

Keywords： three-dimensional space; pedestrian rerecognition; crowd monitoring

1   研究背景

當(dāng)前，目標(biāo)識別、目標(biāo)跟蹤等技術(shù)已廣泛應(yīng)用于視頻監(jiān)控領(lǐng)域。（剩余3938字）