面向地形等高线匹配的三重约束LCSS算法
A Algorithm Based on Triple Constraint LCSS for Terrain Contour Lines Matching

针对地形匹配中数据简化表示以及匹配的问题，提出一种基于等高线特征序列的三重约束LCSS地形匹配方法。首先将基准地形数据用等高线表示，对等高线进行多边形分割逼近，选取分割点作为等高线特征点，然后构造具有平移旋转不变性的弦长夹角特征描述子，对等高线特征点序列进行描述；其次，对实时地形数据以相同的方式构造等高线特征描述子与基准信息匹配；在特征匹配的过程中，针对LCSS算法生成匹配点的外点较多的问题，提出候选集约束、回溯路径同列最小约束、相对位置偏移方差约束的三重约束LCSS方法；最后，采用RANSAC算法对旋转平移参数进行解算，实现地形匹配导航定位。采用秦岭地区ASTER-GTM地形数据验证该地形匹配方法的性能，结果表明文中方法在噪声与几何变换下鲁棒性好，可以显著减少误匹配点数量，提高地形匹配的可靠性，能够有效应用于山区地形匹配导航。
To simplify the representation of terrain and to improve the reliability of terrain matching, a terrain matching method based on contour feature sequence is proposed. In the method, the contour lines are approximated by polygons, selecting the break points as feature points, constructs a translation and rotation invariant feature descriptor on feature points. In view of the false matching problem of contour lines' feature sequence, the candidate set, the backtracking path matching point optimization and the relative position deviation variance constraint method is used to find matching feature point. The performance of the terrain matching method is verified by ASTER-GTM terrain data in Qinling Mountains area. The results show that the proposed method can be applied to the matching of the real-time terrain map and the reference map, robustness to noise and geometric transformations