基于模糊逻辑的多特征融合的SOAMST跟踪算法

SOAMST Tracking Algorithm Based on Fuzzy Logic of Multiple Feature Fusion

针对复杂环境条件下颜色、光照变化和遮挡对目标跟踪算法精度和鲁棒性的影响,论文提出了基于模糊逻辑多特征融合的SOAMST跟踪算法。首先,选择颜色特征和LBP特征对目标进行建模,并根据模糊逻辑算法调整两种特征在计算目标质心位置和权重图像中的权重;其次,目标被遮挡暂时消失时,根据SOAMST算法得到上一帧目标的状态信息,调用粒子滤波算法对目标位置进行预测,可以避免丢失跟踪目标,实现目标连续跟踪。实验表明,论文算法在复杂环境条件下能很好地实现目标跟踪。

In view of the complex environment conditions which are color,illumination changes and occlusion influence target tracking algorithm accuracy and robustness,SOAMST tracking algorithm based on fuzzy logic multiple feature fusion is proposed in this paper.In the first place,a color feature and LBP feature choosen to establish target modeling,and according to the fuzzy logic algorithm the weights of these two characteristics in calculating barycentric position of the target and weight image are adjusted.Then,when the target is occluded and disappeared,the preceding frame of the target state information is obtained,and target positioninvoking particle filter algorithm is forecasted,in this way it can avoid losing track the target,thus realizing continuous target tracking.Experiments show that the algorithm in this paper can well realize target tracking on the complex environment conditions.