基于OTSU多阈值分割算法的激光线扫点云数据表达及精简方法研究  

Research on an expression &reduction method of laser point cloud data based on OTSU multi-threshold segmentation algorithm

作　　者：陈兆学[1] 姚春荣[1] 

机构地区：[1]上海理工大学医疗器械与食品学院,上海200093

出　　处：《光学技术》2016年第5期435-440,共6页Optical Technique

基　　金：上海市教委科研创新项目(13YZ069);上海市自然科学基金项目

摘　　要：激光线扫方式获取的点云数据量庞大,不利于点云数据的存储、处理与分析。为了对激光线扫点云数据进行有效精简,提出了一种基于OTSU多阈值分割算法的激光线扫点云数据表达及精简方法。基于点云数据坐标与图像灰度值的映射,采用OTSU多阈值分割算法进行区域分割,并将分割后的各区域进行边缘提取及细化处理。根据原映射关系将细化的二值图像重新以点云方式表示,即得到精简的点云数据。在保持原有点云数据关键信息完整度的基础上,可有效地精简点云数据。实验结果表明:基于OTSU多阈值分割算法可有效地精简点云数据,同时能够有效地去除扫描过程中的背景干扰数据,具有较大的适用性和实际应用参考价值。The amount of point cloud data obtained based on laser line scan mode is increasingly large. It＇s extremely disadvantageous to the storage, processing and analysis of point cloud data and thus the effective laser point cloud data reduction is needed. An expression ＆ reduction method of laser point cloud data based on OTSU multi-threshold segmenta- tion algorithm is presented. According to the mapping relationship between point cloud data and images gray values, the algorithm of OTSU multi-threshold segmentation is adopted and the edge extraction and refining of segmented regions are performed to acquire the final reduced point cloud data. The method can simplify large amounts of point cloud data effec- tively on the basis of maintaining the integrity of the key information of original point cloud data. The results show that the point cloud data can be simplified effectively based on the method of multi-threshold segmentation algorithm, and at the same time the background interference data in the process of scanning are effectively removed. The method has great applicability and practical application reference values.

关 键 词：图像分割 多阈值 激光点云数据 线扫方式 数据精简 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]