一种camshift算法与brisk特征点相结合的运动目标跟踪方法

A Combination of Camshift Algorithm and Brisk Feature Point for Real Time Moving Target Tracking

传统的camshift算法中,当目标移动过快或背景同目标相似时容易导致目标跟踪丢失,在各种复杂情况下,如何保持跟踪的有效性及连续性成为众多研究人员不断努力的方向。为此,提出了一种brisk特征点检测与camshift算法相结合的目标跟踪方法,以及简化的搜索窗口修正方法,在camshift算法基础上,通过不同帧的目标上brisk特征点匹配来防止局部最大值的出现以及目标移动过快导致的跟踪窗口丢失。实验结果表明:该方法有效地提高了目标跟踪的连续性和稳定性,与其他特征点结合的算法相比有更好的实时性及精确性。

In traditional camshift algorithm, when the target moves too fast or the background has similar color, it will lead to the loss of target easily, and how to maintain the effectiveness and continuity of the tracking becomes the direction for many researchers under various complex situations. Therefore, we proposed a improved camshift algorithm based on combination of eamshift and brisk feature point and a simplified search window correction method, which prevent the emergence of a local maxmum and the loss of tracking window with matching different frames on the goal of your brisk feature point. Experiments show that our method effectively improve the tracking. In combination with other feature points algorithm, it and accuracy and stability of target has better real-time comparison