在处理尺度变化和目标遮挡方面,利用相关滤波器的不同特征进行目标跟踪仍然存在问题.提出了一种基于随机蕨丛检测器的多尺度核相关滤波器算法.该算法将跟踪任务分解为目标尺度估计和位移估计,同时将CN颜色特征和HOG特征进行响应融合,进一步提高了整体跟踪性能.此外,文中训练了一个在线随机蕨分类器,在目标丢失后其能重新获取目标.与KCF,DSST,TLD,MIL,CT共5种算法相比,所提算法不仅能够准确地估计目标状态,而且可以有效处理目标的遮挡问题.%There are still some problems in the aspect of handling scale and object occlusion by using different features of correlation filter to perform object tracking.In this paper,a multi-scale kernel correlation filter algorithm based on random fern detector was proposed.The tracking task was decomposed into the target scale estimation and the translation estimation.At the same time,the CN colour feature and HOG feature were fused in response level to further improve the overall tracking performance of the algorithm.In addition,an online random fern classifier was trained to reobtain the target after the target was lost.By comparing with KCF,DSST,TLD,MIL and CT algorithms,it is proved that the proposed method can accurately estimate target status and effectively deal with the occlusion problem.
展开▼
