摘要
提出基于特征角点的控制点自适应匹配算法：依据图像数据灰度特征自动确定相应阈值,从基准影像上提取控制点,采用动态模板进行不等距搜索,确定潜在目标控制点；构建三角网,利用等角变换判别目标控制点；构建仿射变换方程来筛选目标控制点。该算法无需设定灰度相关系数最大值阈值即可判别目标控制点。选取ASTER和TM两种成像差异显著的图像数据,对该算法进行试验,结果表明,该算法可以准确提取有价值的特征角点,具有较高的控制点识别精度和运算效率,具有较强的适应性和应用价值,较之传统匹配算法有明显优势。
An adaptive point matching algorithm was introduced by this paper. It included following several aspects： The first step was to extract ground control points by an adaptive feature corner extraction method, and the threshold for comer extraction was determined by image gray feature. The second step was to search the sub image unequdistantly with dynamic template, and the purpose was to determine potential control points. The third step was to construct triangle net, to locate the true target points by conformal transform, and traditional method with maximum threshold of gray correlation coefficient was not nseful for this. Moreover, an experiment on the algorithm was performed with an ASTER image and a TM image, and there were some distinct imaging differences of each other. From the result of the experiment, we can conclude that the modified algorithm is superior to the traditional algorithm,for it can extract valuable feature corners, and it has higher accuracy and efficiency than the latter, and it has more adaptability and applicable value.
出处
《宇航学报》
EI
CAS
CSCD
北大核心
2006年第1期45-50,88,共7页
Journal of Astronautics
基金
国家高技术研究发展计划(863计划2003AA131170)中国科学院知识创新工程重大项目(KZCX1-SW-01-02)
关键词
角点提取
控制点匹配
动态模板
等角变换
仿射变换
Corner extraction
Ground control point matching
Dynamic template
Searching uneqnidistantly
Conformal transform
