基于特征异质性优选的面向对象建筑物提取

Object Oriented Building Extraction Based on Feature Heterogeneity Optimization

利用面向对象方法提取遥感影像中的建筑物是当前研究热点之一。针对此方法中特征选择问题,文章基于面向对象的分析方法,充分考虑影像特征与多种地物在空间异质性上的相似程度,提出了一种特征优选方法,进而完成建筑物的提取。具体方法是:首先,计算对象内同质性与对象间异质性,基于均值变点原理确定全局最优尺度;其次,利用地理探测器基于空间异质性相似度对分割对象的光谱、纹理、高度、形状等特征进行优选;最后,根据优选出的特征进行建筑物提取,并辅以邻域高程比对算法进行优化,获得建筑物较完整、可靠的轮廓。结果表明,该方法对建筑物提取更加高效且精度更高。

Using object-oriented method to extract buildings from remote sensing images is one of the current research hotspots.Aiming at the problem of feature selection in this method,based on the object-oriented analysis method,this paper fully considers the similarity between image features and a variety of ground objects in spatial heterogeneity,and proposes a feature optimization method to complete the extraction of buildings.The specific methods are as follows:firstly,calculate the homogeneity within the object and the heterogeneity between objects,and determine the global optimal scale based on the mean change point principle;secondly,use the geographic detector to optimize the spectral,texture,height,shape and other features of the segmented object based on the similarity of spatial heterogeneity;finally,extract the building according to the optimized features,and optimize it with the neighborhood elevation comparison algorithm,so as to obtain a relatively complete and Reliable profile.The results show that this method is more efficient and accurate for building extraction.