密度聚类算法在连续分布点云去噪中的应用

在原始测量获取的点云数据中,除了目标数据外,还有大量的噪声数据。噪声往往无规律地分布在目标物体周围,难以用统一数学模型区分。基于密度的聚类算法将簇定义为密度相连的点的最大集合,能发现任意形状、大小的类簇,将该算法应用在点云去噪中,能将密度分布连续点进行聚类,从中提取出目标点云。

Application of Density-based Clustering Algorithms in Noise Removing of Continuous Distributed Point Clouds

There are lots of noise data in the raw data except the target data. And the noise data always distribute around the target object irregularly, it is impossible to build a math model to make a distinction between the noise data and target data. The cluster is defined as the maximum set of density-collected in Density-based clustering algorithms, it can discover arbitrary shaped or sized cluster. To apply this algorithm it the noise removing of point clouds, can make the continuous distributed points as a cluster, and then extract the target point clouds.