| 一种可见光影像稠密匹配点云的单栋建筑物提取方法 |
| |
| 引用本文： | 杨振发,万刚,曹雪峰,李锋.一种可见光影像稠密匹配点云的单栋建筑物提取方法[J].测绘与空间地理信息,2016(2):8-11. |
| |
| 作者姓名： | 杨振发 万刚 曹雪峰 李锋 |
| |
| 作者单位： | 信息工程大学,河南郑州,450052 |
| |
| 基金项目： | 国家自然科学基金项目(41371384、41401465)；地理信息工程国家重点实验室项目(SKLGIE2014－2－4－1) |
| |
| 摘 要： | 建筑物点云提取是城市快速三维建模的基础。针对城区中建筑物和树木空间距离较近导致建筑物点云误提取的问题,提出一种颜色约束的欧式聚类算法。该方法利用低空拍摄可见光影像进行三维重建、获取点云数据,在建立点云K邻域索引和表面估计的基础上,以曲率最小的点作为欧式聚类的种子点,将点云的RGB值转换成Lab颜色模型,对建筑物点云的聚类提取进行约束。实验表明,该方法可以有效地解决可见光影像匹配点云中建筑物提取时将树木误提取的问题。
|
| 关 键 词： | 可见光影像匹配点云 建筑物单栋提取 欧式聚类算法 颜色约束 Lab颜色模型 |
An Approach for Building Extraction Based on Dense-matching Points Cloud of Visible Image |
| |
| Abstract: | Abstratc:Building extraction of point cloud is the basis of the rapid urban modeling.A problem of mistaken extraction will be arisen when the spatial distance of buildings and trees are too close in city.To solve this problem, a method of European clustering algorithm based on color constrain was proposed.This method regarded minimum curvature points as the European clustering seeds on the basis of establishing point cloud KNN and estimation surface, and constrained the clustering extraction of trees point cloud by converting the RGB values of point cloud to Lab color space, utilizing the point cloud of images matching.The experiments showed that this method could effectively solve the problem of building mistaken extraction based on the visible image matching point cloud. |
| |
| Keywords: | matching point cloud of visible images building extraction European clustering algorithm color constrain Lab color space |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
