利用Zernike矩对心脏超声图进行光流场计算的邻域优化算法

A Local Optimization Algorithm for Computing 2-D Optical Flow Field of Heart Echocardiograph Image Using Zernike Moment

本文针对心脏超声图的特性 ,介绍了一种计算二维光流场的邻域优化算法 ,通过对Zernike矩的正交性、旋转不变性等特性的分析 ,论证了Zernike矩在心脏超声图中作为运动不变特征的优越性 ,从而将传统光流场计算中的亮度不变假设推广到了亮度各阶Zernike矩不变 .另外 ,由于Zernike矩计算了像素点附近其余像素的贡献 ,使得单个像素点引起的误差对特征的影响大大减小 。

This paper presents a local optimization algorithm for computation of 2 D optical flow field aiming at the specialty of heart echocardiograph image.We have analyzed several advantages of Zernike moments such as orthogonality and rotation invariance,and demonstrate that Zernike moments are superior in representing the local motion invariance in heart echocardiograph image.Accordingly,we generalize the premise of traditional Optical Flow Computation that intensity is invariant in the sequence to the invariance of each order Zernike moment.In addition,in virtue of the computation of whole contribution of the neighborhood around each pixel,Zernike moment based method can greatly reduce the influence of a singular error at one pixel on the total representative,therefore improves the robustness of the algorithm.