基于粒子预测的自适应图像匹配目标跟踪算法

在基于自适应图像匹配跟踪算法上研究了让其结合粒子预测及对模板图像更新的综合算法,对非线性、非高斯问题进行了位置预测,针对Hausdorff算法计算量大的问题,采取了更新结合的方法,改变并对自适应图像匹配跟踪算法的匹配跟踪性能进行分析.实验结果表明,纳入粒子预测的自适应跟踪算法既继承了相关跟踪直观实用的特点,有效的克服障碍物的遮挡问题,又具有较高匹配跟踪稳定性.

Adaptive image matching target tracking algorithm based on position estimation

The idea of the adaptive image matching tracking algorithm, then with position estimation and image template renewal and could effectively cope with complicated nonlinear and/or non-Gaussian problems. In order to reduce the amount of calculation of tracking algorithm based on Hausdorff distance, updated with the method of combining is adopted. The adaptive image matching target tracking algorithm based on the matching tracking is presented and the performance of the algorithm is analyzed. The results show that adaptive particle tracking algorithm inherited the relevant practical tracking visual characteristics, the part occlusion is effectively solved and higher stability of the match tracking.