激光与光电子学进展, 2017, 54 (11): 111502, 网络出版: 2017-11-17
一种改进的离散连续能量最小化多目标跟踪 下载: 561次
An Improved Discrete-Continuous Energy Minimization for Multi-Target Tracking
机器视觉 多目标跟踪 能量最小化 身份标签互换 轨迹分段 machine vision multi-target tracking energy minimization identity switch trajectory segmentation
摘要
针对离散-连续能量最小化(DCEM)方法在复杂场景中对轨迹分段或身份标签互换无法有效处理的问题,提出一种改进的DCEM多目标跟踪方法。该方法通过提取被跟踪目标的多特征融合外观向量,利用不同目标间外观特征向量的欧氏距离设计轨迹的外观约束项,处理身份标签互换问题;通过计算相邻时空域内不同轨迹间的运动相似性和外观相似性,设计后处理过程,合并可能为同一轨迹的短轨迹,处理轨迹分段问题。实验结果表明,平均跟踪准确度提高3.6%,平均跟踪精度提高2.5%,并且身份标签互换和轨迹分段情况得到大幅改善,该方法具有更精确更稳定的跟踪能力。
Abstract
Aiming at the problems that the discrete-continuous energy minimization (DCEM) method cannot effectively deal with the trajectory segmentation and identity switch in complex scene, an improved DCEM multi-target tracking method is put forward. In order to solve identity switch, this method extracts the multi-feature fusion appearance vector of the tracked target, and uses the Euclidean distance between appearance vectors from different targets to design the constraint function for trajectories. Then this method designs the post-processing process for trajectory segmentation problem by merging the short tracklets which have high similarity of motion and appearance in adjacent spatial-temporal neighborhood. Experimental results indicate that the average tracking accuracy and average tracking precision can increase by 3.6% and 2.5%,respectively. Furthermore, the problems of trajectory segmentation and identity switch can be greatly improved. The proposed method has better robustness and the tracking accuracy.
张丽娟, 周治平. 一种改进的离散连续能量最小化多目标跟踪[J]. 激光与光电子学进展, 2017, 54(11): 111502. Zhang Lijuan, Zhou Zhiping. An Improved Discrete-Continuous Energy Minimization for Multi-Target Tracking[J]. Laser & Optoelectronics Progress, 2017, 54(11): 111502.