基于色彩局域搜索的彩色目标快速跟踪方法

Algorithm for Fast Tracking of Color Object Based on Color Local-search

提出了一种基于色彩局域搜索的彩色目标快速跟踪方法。该色彩搜索法是利用目标特征色在目标跟踪点附近搜索目标,而无需对整幅图像地行搜索。其图像处理速度不受图像尺度的影响,因而实时跟踪速度大大提高。实验结果得出:该方法的跟踪速度可达到18.45frames/s,跟踪效率为73.8%。

A fast tracking algorithm based on local-searching of object was proposed, in which color feature is used to search the tracked object around tracking position, instead of scanning image row by row. The image processing is independent from image size, so that the tracking speed can be increased efficiently in real-time tracking. Our experimental result shows that by using the proposed algorithm, 18.45 frames can be processed per second at a tracking efficiency up to 73.8%.