复杂地形DSM的地面点识别及DEM提取  

Ground Points Recognition and DEM Extraction Based on DSM in Complex Terrain

作　　者：喻龙华 王雷光 吴楠[1] 廖声熙[1] 张怀清[3] YU Longhua;WANG Leiguang;WU Nan;LIAO Shengxi;ZHANG Huaiqing(Research Institute of Resources Insects, CAF, Kunming 650224, China;College of Forestry, Southwest Forestry University, Kunming 650224 China;Research Institute of Forestry Resource Information Techniques, CAF, Beijing 100091, China)

机构地区：[1]中国林业科学研究院资源昆虫研究所,云南昆明650224 [2]西南林业大学林学院,云南昆明650224 [3]中国林业科学研究院资源信息研究所,北京100091

出　　处：《测绘通报》2018年第5期59-64,共6页Bulletin of Surveying and Mapping

基　　金：国家自然科学基金(41571372);国家林业局948项目(2012-4-71);国家科技部863课题(2012AA12A306)

摘　　要：目前,基于数字地表模型(DSM)提取数字高程模型(DEM)的研究多以平坦地区为研究对象,且精度较低。本文提出了一种基于区域生长的DEM提取算法,该算法以区域生长算法为基础,采用最大类间方差法(OTSU)实现区域生长中种子点、生长准则和终止条件的自适应选择。该方法不仅可从平坦地区和地形复杂的山区的DSM中识别地面点和提取DEM,且能有效解决区域生长算法将地面和地面附着物(本文中地面附着物以高架道路为例)识别为同一类的问题。与附近最小值法进行试验对比,结果表明,本文算法能够较好地提高DEM提取精度,识别地面点的制图精度达90%以上,可靠性和稳定性较强。At present,researches of the extraction of DEM focused on DSM are mostly based on the flat areas,and the precision of extracting DEM is low. Therefore,in this article,a DEM extraction algorithm based on regional growth is proposed and implements ground points recognition and DEM extraction form DSM in complex terrain. The algorithm uses the maximum interclass variance(OTSU） to realize the adaptive selection of seed points,and growth criterion,and termination condition in regional growth.This method identifies the elevated road from the city DSM data to avoid that the regional growth algorithm wrongly divides the elevated road into the ground because of the continuity between the elevated road and the ground.By comparing with the method of local minimum,the results show that the proposed algorithm can improve the accuracy of DEM extraction,and the mapping precision of ground points recognition exceeds 90%,and experimental results demonstrate that the effectiveness of the proposed method is reliable and steady.

关 键 词：DSM DEM 区域生长 复杂地形 高架道路 

分 类 号：P23[天文地球—摄影测量与遥感]