摘要
针对地面三维激光点云包含大量冗余数据,特征信息不明显等问题,提出一种建筑物点云特征线提取方法。该方法以点到其每3个相邻邻域点所构成平面的距离的平均值进行特征点的提取,实验结果表明,该方法受噪声影响小,可清晰完整的提取建筑物特征线。与现有算法提取的结果相比,提取的特征线更加完整,断裂线较少,且提取的复杂汉字轮廓更加清晰,冗余点较少。
In view of redundancy of point cloud data and the shortcoming of unconspicuous features, this paper presents a method of building' s feature lines extraction based on 3-D Terrestrial laser scanning technology. This method extracted building' s feature points by utilizing the average distance of the point to the plane which formed by the three adjacent neighborhood points. As the result showing, this method is little affect.ed by noise point and can extract building' s feature lines clearly and completely. Compared with existing algorithms, the building feature lines extracted by this method is more complete and left less fracture line, besides, the extracted outline of complex Chinese is much clearer and left less redundant points.
出处
《激光杂志》
北大核心
2016年第3期9-11,共3页
Laser Journal
基金
国家自然科学基金项目(51504239)
江苏高校优势学科建设工程项目(SZBF2011-6-B35)
关键词
地面三维激光扫描
建筑物特征线
相邻邻域点
距离平均值
3-D terrestrial laser scanning
building' s feature lines
adjacent neighborhood points
average distance