CAJ | 학술논문

从LiDAR点云中准确地提取地物是LiDAR数据处理过程中的一项关键工作。针对LiDAR点云数据在滤波分类和点云精确提取分类方面的不足，提出了一种新的影像辅助下LiDAR点云地物提取的方法。首先，对LiDAR点云按高程值重采样生成高程投影影像并自动分割，同时从高分辨率正射影像提取特征影像，然后将特征影像和由LiDAR生成的分割影像进行集成分析，利用影像提取结果辅助LiDAR点云错分、漏分纠正和点云精确分类处理，最后通过试验分析，验证了该方法的可行性与有效性。
종LiDAR점운중준학지제취지물시LiDAR수거처리과정중적일항관건공작。침대LiDAR점운수거재려파분류화점운정학제취분류방면적불족，제출료일충신적영상보조하LiDAR점운지물제취적방법。수선，대LiDAR점운안고정치중채양생성고정투영영상병자동분할，동시종고분변솔정사영상제취특정영상，연후장특정영상화유LiDAR생성적분할영상진행집성분석，이용영상제취결과보조LiDAR점운착분、루분규정화점운정학분류처리，최후통과시험분석，험증료해방법적가행성여유효성。
Accurately capturing geo-spatial features from LiDAR Point clouds is a key work in the process of LiDAR data processing. In view of the deficiency of LiDAR point cloud data classification in filtering and accurate capturing geo-spatial features,this paper presents a method for image-assisted capturing of geo-spatial features from LiDAR point clouds. First,according to the elevation value of LiDAR point clouds, resample and generate the elevation projection image,and automatically segment it. In the meanwhile,capture the features from the high resolution DOM. Then,conduct integrated analysis of the images of the features and segmented images,and process the wrong,leakage,precise classification of the LiDAR point clouds using the result of image extraction. The feasibility and effectiveness of the method are verified with experiments.