集成MSER和SIFT特征的遥感影像自动配准算法  

Remote Sensing Image Automatic Registration Algorithm of Integrated MSER and SIFT Features

作　　者：王晓华[1,2] 邓喀中[2] 杨化超[2] 

机构地区：[1]河南理工大学测绘与国土信息工程学院,河南焦作454000 [2]中国矿业大学江苏省资源环境信息工程重点实验室,江苏徐州221116

出　　处：《光电工程》2013年第12期31-38,共8页Opto-Electronic Engineering

基　　金：中央高校基本科研业务费专项资金资助(2010QNA21);国土环境与灾害监测国家测绘局重点实验室开放基金资助项目(LEDM2011B07)

摘　　要：针对倾斜的遥感影像配准困难问题,提出一种基于集成最大极值稳定区域(MSER)和尺度不变特征转换(SIFT)的互补不变特征的自动影像配准算法。该算法首先应用目前公认的具有最佳仿射不变性的MSER特征区域进行影像的粗匹配,初步校正影像的空间形变。然后在粗匹配基础上采用匹配能力较强的SIFT描述子与仿射不变矩描述子相结合,进行精匹配。通过以上两步匹配,可以提高遥感影像配准精度,尤其对倾斜影像效果更明显。最后采用倾斜的无人机(UAV)影像进行试验,并与SIFT配准算法比较。结果表明,本文算法在仿射不变性和匹配正确率方面均优于SIFT配准方法。An image matching approach which integrates Maximally Stable Extremal Regions （MSER, Maximally Stable Extremal Regions） and Scale Invariant Feature Transformation （SIFT, Scale Invariant Feature Transformation） complementary invariant feature automatically is proposed for the flit Remote Sensing image registration. Firstly, the images are coarsely matched by applying currently recognized as the best affine invariant MSER features, and the large deformation images are corrected initially. Then the images are fine matched by the matching ability of the SIFT descriptor joint the moments based on the coarse matching. The remote sensing image matching accuracy is improved through the above two steps, especially, the more pronounced effect on the large tilt images. Finally, the UAV（Unmanned Aerial Vehicle） image experiments show that this algorithm is more effective than SIFT algorithm in the affine invariant and matching the correct rate

关 键 词：影像配准 互补不变特征 MSER特征 SIFT特征 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]