低空机载LiDAR水面点云自适应分类算法研究

Research on self-adaptive classification algorithm of water point clouds by low-altitude airborne LiDAR

水陆点云分类是DTM生成、河流岸线提取等低空机载LiDAR应用领域面临的新问题,然而在岸滩等复杂扫描场景中,水陆点云的准确分类是一个公认的难题。在分析目前点云分类存在的缺陷基础上,提出了一种多元特征统计的自适应水陆LiDAR点云分类算法,该算法通过分析低空机载LiDAR水面点云的特点,针对性地设计了点云坡度、密度特征描述子;引入贝叶斯定理,建立了高程、坡度、密度隶属度函数;通过水、陆独立样本的t检验,确定隶属度函数的自适应权重;最终得到一个多元特征统计的分类模型,并基于训练样本的概率密度统计,确定了模型的自适应分类阈值。典型应用实例表明,在存在岸滩、内陆平地等复杂地形条件下,新算法都能达到99%以上的水陆点云分类精度。

Water and land point clouds classification is a new issue for the application of low-attitude airborne LiDAR,such as DTM generation and river shoreline extraction.However,for some complex landscapes,distinguishing the water points from the land points is difficult.An self-adaptive classification algorithm of water LiDAR point clouds by multivariate feature statistics is proposed in this paper.Firstly,the operators of local terrain slope and points density are adopted according to the characteristics of low-altitude airborne LiDAR water point clouds.Then,Bayes′theorem is used to construct membership functions of the elevation,slope and density.Then,the adaptive weights of the individual membership functions are determined according to the t-test of the independent-samples of water and land points.Finally,a classification model based on multivariate feature statistics is obtained,and the adaptive classification threshold of the model is determined by the probability density of the training samples.Typical experiments indicate that water classication accuracies higher than 99%can be obtained by this algorithm,even in complex landscapes with mudflat and inland plain.