| 85 | 0 | 135 |
| 下载次数 | 被引频次 | 阅读次数 |
提出一种新型双时特征融合的目标跟踪算法XT-SORT,旨在解决遮挡、远距离检测及复杂场景下的跟踪挑战。该方法在YOLOv10架构中引入多尺度特征提取全局空间注意力机制模块，有效提升小目标和远距离目标的检测精度，同时引入双时特征融合模块，以增强目标重识别能力，提高遮挡和快速运动情况下的跟踪表现。实验结果表明，XT-SORT在MOT17中MOTA指标可达到77.8,特别是在遮挡、远距离目标检测和多目标场景中表现突出，为复杂环境下的目标跟踪提供了高效、精准的解决方案。
Abstract：This paper proposes a novel bitemporal feature fusion-based object tracking algorithm XT-SORT,aiming to address the challenges of tracking in scenarios with occlusion, long-range detection, and complex environments.This method introduces a multi-scale feature extraction and global spatial attention mechanism module in the YOLOv10 architecture, effectively enhancing the detection accuracy of small and long-range targets.Meanwhile, it incorporates a dual-time feature fusion module to strengthen the target re-identification capability and improve tracking performance under occlusion and rapid movement conditions.Experimental results show that XT-SORT achieves a MOTA score of 77.8 in MOT17,particularly excelling in scenarios with occlusion, long-range target detection, and multiple targets, providing an efficient and precise solution for object tracking in complex environments.
[1] REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once:Unified,real-time object detection[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Las Vegas,NV,USA:IEEE Computer Society,2016:779-788.
[2] REDMON J,FARHADI A.YOLO9000:better,faster,stronger[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Honolulu,HI,USA:IEEE Computer Society,2017:6 517-6 525.
[3] REN S Q,HE K M,GIRSHICK R,et al.Faster R-CNN:towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,IEEE Computer Society,2017,39(6):1 137-1 149.
[4] LIU W,ANGUELOV D,ERHAN D,et al.SSD:Single shot MultiBox detector[M]//Computer Vision-ECCV 2016.Cham:Springer International Publishing,2016:21-37.
[5] DALAL N,TRIGGS B.Histograms of oriented gradients for human detection[C]//2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR′05).San Diego,CA,USA:IEEE Computer Society,2005:886-893.
[6] DOSOVITSKIY A,BEYER L,KOLESNIKOV A,et al.An image is worth 16x16 words:transformers for image recognition at scale[EB/OL].(2021-06-03)[2025-03-29].https://arxiv.org/abs/2010.11929.
[7] CARION N,MASSA F,SYNNAEVE G,et al.End-to-end object detection with transformers[M]//Computer Vision-ECCV 2020.Cham:Springer International Publishing,2020:213-229.
[8] ZHOU X Y,WANG D Q,KR?HENBüHL P.Objects as points[EB/OL].(2019-04-16) [2025-02-27].https://arxiv.org/abs/1904.07850v2.
[9] KUHN H W.The Hungarian method for the assignment problem[J].Naval Research Logistics Quarterly,1955,2(1/2):83-97.
[10] KALMAN R E.A new approach to linear filtering and predictionproblems[J].Journal of Basic Engineering,1960,82(1):35-45.
[11] WANG Z D,ZHENG L,LIU Y X,et al.Towards real-time multi-object tracking[M]//Computer Vision-ECCV 2020.Cham:Springer International Publishing,2020:107-122.
[12] WOJKE N,BEWLEY A,PAULUS D.Simple online and realtime tracking with a deep association metric[C]//2017 IEEE International Conference on Image Processing (ICIP),2017:3 645-3 649.
[13] ZHANG Y F,WANG C Y,WANG X G,et al.FairMOT:on the fairness of detection and re-identification in multiple object tracking[J].International Journal of Computer Vision,2021,129(11):3 069-3 087.
[14] SUN P Z,CAO J K,JIANG Y,et al.TransTrack:multiple object tracking with transformer[EB/OL].(2021-05-04) [2025-02-27].https://arxiv.org/abs/2012.15460v2.
[15] ZENG F G,DONG B,ZHANG Y A,et al.MOTR:end-to-end multiple-object tracking with transformer[M]//Computer Vision-ECCV 2022.Cham:Springer,2022:659-675.
[16] BEWLEY A,GE Z Y,OTT L,et al.Simple online and realtime tracking[C]//2016 IEEE International Conference on Image Processing (ICIP),2016:3 464-3 468.
[17] BERGMANN P,MEINHARDT T,LEAL-TAIXE L.Tracking without bells and whistles[C]//2019 IEEE/CVF International Conference on Computer Vision (ICCV),2019:941-951.
[18] ZHANG Y F,SUN P Z,JIANG Y,et al.ByteTrack:multi-object tracking by associating every detection box[M]//Computer Vision-ECCV 2022.Cham:Springer,2022:1-21.
[19] CAO J K,PANG J M,WENG X S,et al.Observation-centric SORT:rethinking SORT for robust multi-object tracking[EB/OL].(2023-03-16)[2025-02-27].https://arxiv.org/abs/2203.14360v3.
[20] QIN Z,ZHOU S P,WANG L,et al.MotionTrack:learning robust short-term and long-term motions for multi-object tracking[EB/OL].(2023-04-17)[2025-02-27].https://arxiv.org/abs/2303.10404v2.
[21] HAN X D,OISHI N,TIAN Y Y,et al.ETTrack:enhanced temporal motion predictor for multi-object tracking[J].Applied Intelligence,2024,55(1):33.
[22] ZHANG G Y,WANG C B,GAO W.Pedestrian multiobject tracking algorithm with anti-occlusion[J].CAAI Transactions on Intelligent Systems,2024,19(5):1 248-1 256.
基本信息:
DOI：10.19724/j.cnki.jmju.2025.05.004
中图分类号:TP391.41
引用信息:
[1]石家劲,林贵敏,尹威.双时特征融合的目标跟踪算法[J].闽江学院学报,2025,46(05):31-40.DOI:10.19724/j.cnki.jmju.2025.05.004.
基金信息:
福建省自然科学基金项目(2023J011401); 福建省本科高校教育教学研究项目(重大项目)(FBJY20230095)
2025-03-29
2025
2025-09-23
2025
1
2025-09-26
2025-09-26
2025-09-26