数字图像辅助激光点云特征提取  

Digital Image Assisted Point Cloud Feature Extraction

作　　者：周春艳[1] 罗敏[1] 邹峥嵘[2] 

机构地区：[1]中南大学信息科学与工程学院,湖南长沙410083 [2]中南大学地球科学与信息物理学院,湖南长沙410083

出　　处：《微计算机信息》2011年第10期140-141,131,共3页Control & Automation

摘　　要：提出了一种新颖的激光点云特征提取方法。将点云数据中的点映射到二维图像中的像素,然后使用数字图像处理算法从二维图像中提取感兴趣的特征,根据对应关系得到图像中的特征在点云中的对应点集,对点集进行拟合以剔除噪声点获得更加精确的特征点集。实验表明,此方法能够快速和比较精确地提取点云特征。A novel method for feature extraction from point cloud is proposed.Each point in the point cloud is corresponded to pixel in the digital image,then features of interest are extracted from the digital image by using existed image processing algorithms,according to relationship between the digital image and the point cloud,features in digital image can find their corresponding points in point cloud,and these points are the very points set of features.Points set are fitted to obtain more accurate feature points set.It is proven that this method can extract the point cloud features quickly and more precisely.

关 键 词：三维激光扫描 配准 点云特征 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]