基于运动矢量场的双迭代全局运动估计方法  

Double-iteration algorithm for motion field based global motion estimation

作　　者：陈韩锋[1] 戚飞虎[1] 

机构地区：[1]上海交通大学计算机科学与工程系,上海200030

出　　处：《通信学报》2004年第6期126-131,共6页Journal on Communications

基　　金：国家自然科学基金资助项目(60072029)

摘　　要：提出了一种基于运动矢量场的双迭代全局运动估计方法,该方法用两个最小二乘迭代过程来去除局部运动区域对全局运动估计的干扰。第一个迭代使用一个递减的百分比阈值来排除局部运动区域;第二个迭代过程使用一个固定的绝对阈值来检测完整的全局运动区域并估计运动模型参数。实验结果表明双迭代法进行全局运动估计的结果更加准确和稳定。A double-iteration algorithm with variable thresholds is proposed to detect the global motion regions (GMR) and estimate parameters of the global motion model (PGMM). The proposed algorithm discriminates local motions from global motion by two iterative processes. In the first iterative process, a degressive threshold is introduced to eliminate the disturbance of local motion. Then, the GMR is detected and the final PGMM is calculated on the GMR in the second iterative process where a static threshold is used. Experiments demonstrate that the proposed algorithm is more accurate and robust in getting the GMR and PGMM than many previous algorithms.

关 键 词：图像处理 全局运动估计 摄像机运动模型 双迭代 

分 类 号：TN391[电子电信—物理电子学]