基于最小二乘法的激光雷达数据滤波方法  

A Filtering Algorithm of Airborne Li DAR Points Cloud Based on Least Square

作　　者：赵慧珍[1] 程英蕾[1] 屈亚运 

机构地区：[1]空军工程大学信息与导航学院,西安710077

出　　处：《科学技术与工程》2014年第33期234-239,共6页Science Technology and Engineering

基　　金：国家自然科学基金项目(61175029);陕西省自然科学基金(2011JM8015)资助

摘　　要：最小二乘算法在"等权"的条件下进行,不具有先验知识,不适用地形复杂的区域。针对此问题,提出了一种改进的基于最小二乘的激光雷达(light detection and ranging,Li DAR)数据滤波方法。首先对离散点高程进行腐蚀操作,将得到的高程与原始高程相比,选取小于阈值条件的点作为种子点。其次根据腐蚀前后的高程差值赋予点不同的权重,利用最小二乘原理拟合这些种子点,计算所有点的拟合值与实际高程的差;特定阈值比较并将小于阈值的点加入种子点,迭代至地面点集不再变化。利用国际摄影测重与遥感学会(international society for photogrammetry and remote sensing,ISPRS)提供的数据进行实验,滤波总误差小于经典算法。实验结果表明算法能够较好的滤除地物,适用于地形复杂的区域。All points have the same weight at first so the algorithm based on least square does not have the priori-knowledge,and cannot be applied to complex scenes. In order to solve this problem,an improved filtering algorithm of airborne Li DAR data based on least square was proposed. Firstly,process the height of points with erosion operator,a point is classified into seed set if its difference between the original height and the height after erosion operation is more than a certain threshold. And then different weight given to a point according to the difference of height. These seeds are fited into curved surface by the least square method. The distance is compared to the surface of other points with a certain elevation difference threshold. And the point into seed set if the distance is smaller than the threshold is put out. Fit a curved surface and iterate until there is no seed generated. Data provided by ISPRS is used for experiment and the total error of filtering is smaller than classic algorithms. The result of experiment showed that the method could have a good effect on filtering.

关 键 词：激光雷达 滤波 腐蚀 最小二乘 

分 类 号：TN957.513[电子电信—信号与信息处理]