同名点及高程平面约束的航空影像直线匹配算法

针对直线匹配的难点问题及匹配约束的有效性,提出了同名点及高程平面约束的航空影像直线匹配算法。该算法在边缘点匹配结果和直线提取结果的基础上,首先利用直线邻域内的同名点确定候选直线及直线投影平面的高程值,再结合物方和像方相似性约束确定同名直线;然后根据直线索引对"一配多"的匹配结果进行整合,并对结果中的多直线进行合并,得到"一对一"的同名直线;最后利用"像方-物方-像方"的映射模式确定同名直线的同名端点。论文选取典型纹理特征的航空影像进行直线匹配试验,结果表明,本文算法能获得可靠的直线匹配结果。

Line Matching Algorithm for Aerial Image Based on Corresponding Points and Z-plane Constraints

Aiming at the problem of line matching and the effectiveness of matching constraints, this paper presents one new line matching algorithm based on corresponding points and Z-plane constraints. Based on the edge points matching results and the line extraction results, the algorithm firstly determines the candidate lines by corresponding points in neighborhoods of lines, also the elevation values of line projection planes, and then obtains the corresponding lines with the similarity constraint between object space and image space. Secondly, it integrates the matching results of “one-to-many” relationship according to the line indexes and merges those into one line, and then translates the matching results into “one-to-one” relationship. Finally, it determines the corresponding endpoints of corresponding lines by using “image-object-image” mapping mode. In the experiments, this paper adopts digital aerial images with classic texture features for line matching, and the experiment results demonstrate that the proposed algorithm in this paper can obtain reliable line matching results.