Multiple Object Tracking (MOT) aims to detect and track multiple targets across consecutive video frames while preserving consistent object identities. While appearance-based approaches have achieved notable success, they often struggle in challenging conditions such as occlusions, motion blur, and the presence of visually similar objects, resulting in identity switches and fragmented trajectories. To address these limitations, we propose Motion-Perception Multi-Object Tracking (MPMOT), a motion-aware tracking framework that emphasizes robust motion modeling and adaptive association. MPMOT incorporates three core components: (1) a Gain Kalman Filter (GKF) that adaptively adjusts detection noise based on confidence scores, stabilizing motion prediction during uncertain observations; (2) an Adaptive Cost Matrix (ACM) that dynamically fuses motion and appearance cues during track–detection association, improving robustness under ambiguity; and (3) a Global Connection Model (GCM) that reconnects fragmented tracklets by modeling spatio-temporal consistency. Extensive experiments on the MOT16, MOT17, and MOT20 benchmarks demonstrate that MPMOT consistently outperforms state-of-the-art trackers, achieving IDF1 scores of 72.8% and 72.6% on MOT16 and MOT17, respectively, surpassing the widely used FairMOT baseline by 1.1% and 1.3%. Additionally, rigorous statistical validation through post hoc analysis confirms that MPMOT’s improvements in tracking accuracy and identity preservation are statistically significant across all datasets. MPMOT delivers these gains while maintaining real-time performance, making it a scalable and reliable solution for multi-object tracking in dynamic and crowded environments.

When the multi-object tracking (MOT) algorithm confronts complex scenarios such as target occlusion and blurring, the trajectory missing and identity switching problems frequently occur. To address this issue, a trajectory association strategy for MOT is presented. First, using the similarity relationship between two frames, the similarity matrix is introduced to predict a more trustworthy offset vector and lower the number of Identity Switch. Then, a trajectory reconnection strategy is provided in order to retain a portion of the low-score detection frames due to occlusion and to perform secondary matching on the low-score detection frames in order to reduce missing trajectories. On a typical multi-object tracking dataset MOT17, experimental comparison study is performed with several known advanced algorithms, and MOTA and IDF1 achieve 67.9% and 67.8%, respectively. The experimental results show that the proposed algorithm better solves the trajectory missing and identity switching problems, and achieves excellent tracking performance in complex scenarios.