基于改进型随机抽样一致法的图像配准算法  

Image Registration Algorithm Based on Improved Random Sample Consensus Method

作　　者：刘江明[1] 姚素英[1] 史再峰[1] 庞科[1] 

机构地区：[1]天津大学电子信息工程学院,天津300072

出　　处：《南开大学学报（自然科学版）》2016年第1期23-29,共7页Acta Scientiarum Naturalium Universitatis Nankaiensis

基　　金：国家高技术研究发展计划("863"计划)(2012AA012705);国际科技合作计划(2012DFB10170)

摘　　要：为了提高相似变换图像配准的速度和精度,提出了1种基于改进型随机抽样一致法的图像配准算法.在利用Harris角点检测提取待配准图像的特征点以及利用归一化互相关粗匹配后,采用改进的随机抽样一致法进行快速精准的变换模型估计.算法采用图像相似变换的简化配准模型,利用相似特征3角形进行快速模型预检验,并使用最大欧氏距离法提高计算数据的可靠性.实验结果表明,该算法在具有较高计算精度和鲁棒性的情况下,大幅减少了运算量,提高了变换模型的计算速度.In order to improve the speed and accuracy of similarity transformation image registration,an image registration algorithm based on improved random sample consensus method was proposed. After using the Harris corner detection to extract feature points for registration and using the normalized crosscorrelation to complete coarse matching, the improved algorithm was used to estimate transformation model fast and accurately. The algorithm adopted simplified registration model of image similarity transformation,and quickly pre- evaluated the models using the similar feature triangles. The largest Euclidean distance method was also used to improve the reliability of the calculated data. The experimental results show that the improved algorithm has higher calculation accuracy and robustness and significantly reduces the computational complexity, raising the speed of computing transformation model.

关 键 词：图像配准 随机抽样一致 相似变换 相似特征3角形 预检验 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]