复杂环境基于多信息融合的车辆跟踪方法

平交路口复杂环境下基于视觉的车辆跟踪容易受到如车辆在图像上投影的尺 度变化，车辆的排队与消散过程中邻近车辆间的遮挡及分离等因素的影响.针对该问题， 本文提出了一种利用局部特征增强的Mean-shift 改进算法，利用SIFT 特征点对尺度、旋 转变化鲁棒的特性，将其与基于跟踪区域颜色特征的跟踪方法相融合实现车辆跟踪，较 好地解决了在车辆尺度、运动方向变化，以及遮挡情况下的跟踪问题.同时通过引入跟踪 车辆分离的判定条件，结合特征点聚类算法解决了相邻车辆发生分离时的判断及跟踪问 题.实验结果表明，在多种交通场景的车辆跟踪过程中，本文提出的算法有较好的鲁棒性, 定位结果更加精确.

Multiple Information Fusion-based Vehicle Tracking in Complex Environment

The vehicle tracking at intersection environment is affected by vehicle scale change, occlusion, vehicle queue and dissipation. To solve the problem, a reinforcement algorithm is proposed by using the rotation robustness and scale-invariant characteristics of SIFT feature, combing with region color histogram to track vehicle in scale change and occlusion situation. This paper also establishes criteria for vehicle separation and integrates clustering algorithm to detect and track departing vehicle. The experiment result shows that compare with traditional algorithm, the proposed algorithm provide more robust and more accurate result in multiple traffic scenes.