基于遗传算法的非刚性体光流场估计方法

摘要

In this paper a genetic algorithm (GA) based method for optical flow estimation of nonrigid object is presented. The nonrigid body is divided into several infinitesimal patches in which the pixels move together. The motion parameters of the patches are optimally estimated using GA method. The application of this method in motion estimation of biological tissues is also analyzed. The experimental results show that this method has the advantages of robustness,global optimality and high convergent speed and it is feasible for optical flow estimation of nonrigid body.%本文提出了一种可用于非刚性体的基于遗传算法(Genetic Algorithm，GA)的光流场估计方法，将非刚性体分割为共同运动的微元，利用遗传算法对各运动参数进行优化估计。随后，研究了该方法在生物组织运动估计中的应用。实验结果表明：本方法具有鲁棒性好、全局搜索能力强、收敛速度快和可用于非刚性体运动估计等优点。