顾及邻域相似性的面要素匹配方法  

Matching method for areal feature considering Context-dependent similarity

作　　者：焦洋洋[1] 王卉[2] 翟仁健[1] 

机构地区：[1]信息工程大学地理空间信息学院,河南郑州450052 [2]国防信息学院,湖北武汉430010

出　　处：《辽宁工程技术大学学报（自然科学版）》2013年第7期947-952,共6页Journal of Liaoning Technical University (Natural Science)

基　　金：国家自然科学基金资助项目(41171354;41101362;41171305);国家"863"计划资助项目(2009AA12Z305)

摘　　要：针对现有面要素匹配方法中只注重个体相似性度量、对空间关系相似性不够重视等问题,采用顾及邻域相似性的面要素匹配方法,进一步完善传统的基于几何相似性评价的匹配方法,使用相似矩阵法确定匹配关系;考虑要素空间关系的相似性,与几何相似性结合得到顾及邻域相似性的评价模型.研究结果表明:该方法能够有效地解决待匹配要素位置偏移降低匹配精度等问题,进一步提高了匹配的速度和正确率.Aiming at the problems that the existing matching methods emphasize on geometry similarity,and not enough on spatial relations similarity,this paper put forwards a matching method for areal feature considering Context-dependent similarity.This method improved the traditional matching method based on geometry similarity evaluation,and then provide a similar matrix method to ensure matching relationship.This paper considered the similarity of spatial relationship,and then set up the evaluation model considering context-dependent similarity by combining the geometry similarity.The results of study show that this method can effectively solve the problem of the low matching precision caused by coordinate offset,etc.,and further enhance the speed and accuracy of the matching.

关 键 词：面要素 匹配 个体几何 邻域空间结构 相似性评价 相似矩阵 更新 植被 

分 类 号：N5[自然科学总论]