《计算机应用》唯一官方网站 ›› 2026, Vol. 46 ›› Issue (9): 2776-2786.DOI: 10.11772/j.issn.1001-9081.2025081053

• 人工智能 • 上一篇    

时空建模与层次化特征增强的行人重识别算法

杨定礼(), 卫元芳, 胡文瑞, 孔力杨, 于银山   

  1. 淮安大学 电子信息工程学院,江苏 淮安 223003
  • 收稿日期:2025-09-11 修回日期:2025-12-18 接受日期:2026-01-04 发布日期:2026-02-12 出版日期:2026-09-10
  • 通讯作者: 杨定礼
  • 作者简介:杨定礼(1973—),男,淮安金湖人,副教授,硕士,CCF会员,主要研究方向:深度学习、图像信号处理
    卫元芳(1998—),女,安徽六安人,硕士研究生,主要研究方向:深度学习、行人重识别
    胡文瑞(1999—),男,安徽淮南人,硕士研究生,主要研究方向:深度学习、车道线检测
    孔力杨(2001—),男,江苏盐城人,硕士研究生,主要研究方向:目标检测、图像增强
    于银山(1987—),男,山东聊城人,副教授,博士,主要研究方向:图像处理、机器视觉。
  • 基金资助:
    国家自然科学基金资助项目(61801188)

Spatio-temporal modeling and hierarchical feature enhancement for person re-identification

Dingli YANG(), Yuanfang WEI, Wenrui HU, Liyang KONG, Yinshan YU   

  1. Faculty of Electronic Information Engineering,Huai’an University,Huai’an Jiangsu 223003,China
  • Received:2025-09-11 Revised:2025-12-18 Accepted:2026-01-04 Online:2026-02-12 Published:2026-09-10
  • Contact: Dingli YANG
  • About author:YANG Dingli, born in 1973, M. S., associate professor. His research interests include deep learning, image signal processing.
    WEI Yuanfang, born in 1998, M. S. candidate. Her research interests include deep learning, person re-identification.
    HU Wenrui, born in 1999, M. S. candidate. His research interests include deep learning, lane detection.
    KONG Liyang, born in 2001, M. S. candidate. His research interests include object detection, image enhancement.
    YU Yinshan, born in 1987, Ph. D., associate professor. His research interests include image processing, machine vision.
  • Supported by:
    National Natural Science Foundation of China(61801188)

摘要:

针对遮挡、视角变化及姿态变化等复杂场景下的行人重识别(Re-ID)匹配困难的问题,提出一种时空建模与层次化特征增强的Re-ID算法,通过三阶段渐进式的特征优化框架,实现全局一致性与局部判别性的协同提升。首先,在骨干网络提取外观特征后,引入双池化时序注意力机制,该机制结合全局平均池化与时间平均池化捕获序列特征的互补信息,并通过通道与空间交互进行时空依赖建模,从而突出运动相关特征并缓解遮挡导致的局部信息缺失;其次,针对人体部位特征分布不均的问题,构建柔性特征融合模块,通过可学习权重自适应地聚合多部位特征,抑制遮挡噪声并增强判别性局部特征,从而获得全局与局部层次化表示;最后,在分类层前设计置信度校准网络,通过残差学习优化身份预测的置信度分布,提升跨摄像头检索精度。在Market-1501公开数据集上进行系统评估的结果表明,所提算法的平均精度均值(mAP)达到93.2%,Rank-1达到97.4%,相较于基线算法BPBReIDHR(Body Part-Based ReID model),分别提升了3.8和1.7个百分点。所提算法通过时空建模与层次化特征增强的结合能显著提升复杂场景下的Re-ID性能。

关键词: 行人重识别, 时空建模, 柔性特征融合, 特征增强, 遮挡

Abstract:

To address the matching difficulties in person Re-IDentification (Re-ID) under complex scenes such as occlusion, viewpoint changes, and pose variations, an Re-ID algorithm based on spatio-temporal modeling and hierarchical feature enhancement was proposed. In the algorithm, a collaborative improvement of global consistency and local discriminability was achieved through a three-stage progressive feature optimization framework. First, after extracting appearance features by the backbone network, a dual-pooling temporal attention mechanism was introduced. In the mechanism, complementary information of sequence features was captured by combining global average pooling and temporal average pooling, and spatio-temporal dependency modeling was performed through channel and spatial interaction, so that the motion-related features were highlighted and the local information loss caused by occlusion was alleviated. Second, to address the problem of uneven distribution of human body part features, a flexible feature fusion module was constructed, so that the multi-part features were aggregated through learnable weights adaptively, occlusion noise was suppressed, and discriminative local features were enhanced, thereby obtaining the global and local hierarchical representations. Finally, a confidence correction network was designed before the classification layer, so that the identity prediction confidence distribution was optimized through residual learning, and the cross-camera retrieval accuracy was improved. The results of systematic evaluation on the Market-1501 public dataset show that the proposed algorithm achieves an mAP of 93.2% and a Rank-1 of 97.4%, which are increased by 3.8 and 1.7 percentage points compared with the baseline algorithm Body Part-Based ReID model (BPBReIDHR). It can be seen that the proposed algorithm improves the Re-ID performance in complex scenes significantly by the combination of spatio-temporal modeling and hierarchical feature enhancement.

Key words: Person Re-Identification (Re-ID), spatio-temporal modeling, flexible feature fusion, feature enhancement, occlusion

中图分类号: