作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2009, Vol. 35 ›› Issue (13): 202-204. doi: 10.3969/j.issn.1000-3428.2009.13.070

• 人工智能及识别技术 • 上一篇    下一篇

基于复小波的彩色光流估计

张 磊,项学智,赵春晖,郭黎利   

  1. (哈尔滨工程大学信息与通信工程学院,哈尔滨 150001)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2009-07-05 发布日期:2009-07-05

Color Optical Flow Estimation Based on Complex Value Wavelet

ZHANG Lei, XIANG Xue-zhi, ZHAO Chun-hui, GUO Li-li   

  1. (School of Information and Communication Engineering, Harbin Engineering University, Harbin 150001)
  • Received:1900-01-01 Revised:1900-01-01 Online:2009-07-05 Published:2009-07-05

摘要: 利用色彩恒常性假设,将RGB色彩分量光流基本方程映射至小波域,通过复小波构建的过定线性方程组进行多分辨率光流估计,避免实小波相位震荡带来的光流求解误差,并利用各颜色分量组成的线性方程组的系数矩阵条件数判断求解的可靠性,选择各颜色分量中最可靠的解最为该点最终的光流解。数值实验结果证明了该算法的有效性。

关键词: 光流, 小波, 彩色图像, 运动估计

Abstract: Using assumption of color invariance, basic optical flow equation based RGB color components is mapped to the field of wavelet. Complex value wavelet is used to produce over-determined system of equations and computes the optical flow in multi-resolution scheme. Complex value wavelet can also be used to reduce the effect of phase vibrating. The condition number of the linear system of equations composed by RGB color components is used to judge the reliability of the optical flow result. The reliable result is selected to be the final optical flow component value. Results of numeric experiments show that the method is efficient.

Key words: optical flow, wavelet, color image, motion estimation

中图分类号: