基于灰度共生的多线索目标联合优化跟踪

Object Joint Optimization Tracking Based on Gray-Level Co-Occurrence and Multi-Clues

为了提高跟踪算法对多种目标表观变化场景的自适应能力与跟踪精度,提出一种基于灰度共生的多线索目标联合优化跟踪算法。该算法首先提取目标灰度信息,通过灰度共生的高区分度特征对目标进行二元超分描述,结合三阶张量理论融合目标区域的多视图信息,建立起目标的三维在线表观模型,然后利用线性空间理论对表观模型进行双线性展开,通过双线性空间的增量学习更新,降低模型更新时的运算量。通过二级联合跟踪机制对跟踪估计进行动态调整,以避免误差累积出现跟踪漂移。与典型算法进行多场景试验对比,表明该算法能有效地应对多种复杂场景下的运动目标跟踪。

In order to improve the stability of the object tracking under different conditions, an object tracking algorithm is proposed. First, the algorithm extracts the gray information of target to describe the two high discrimination features of target by gray-level co-occurrence matrix, the dynamic information about target changing is fused by third-order tensor theory, the three-dimensional online appearance model of the object is constructed. Then, bilinear space theory is used to expand the appearance model, implement the incremental learning of model updating, and reduce the computation of the model updating. The secondary combined stable tracking of object is achieved by dynamic matching of two observation models. Experimental results indicate that the proposed algorithm can effectively deal with the moving object tracking on a variety of challenging scenes.