多解析度匹配算法在高速运动物体跟踪中的应用

MULTI-RESOLUTION TEMPLATE-BASED HIGH SPEED OBJECT ANALYSIS

本文提出了一种使用多解析度匹配算法来跟踪高速运动物体的方法 ,并且采用了Kalman滤波器和遗传算法构建了整个的分析框架。通过对高速车辆碰撞试验的结果分析 ,可以看出本算法框架在点跟踪处理中 ,在精确度和容错性方面都能够得到较为满意的结果。

The article provides a method to track high-speed objects by using multi-resolution template matching algorithm.The framework of analyzing tracked features also employs Kalman filter and Genetic algorithm.The analysis results demonstrate that tracking features,using this algorithm framework,is not only accurate but also robust.