Improved Euclidean clustering point cloud segmentation algorithm based on curvature constraint

In response to the problem that the traditional segmentation algorithm is not ideal for segmenting point cloud data in parts with large changes in geometric features and complex shapes, a curvature constraint-based Euclidean clustering algorithm is proposed. This method first calculates the curvature information corresponding to each point according to the spatial coordinate information of the point cloud, then redefines the improved Euclidean distance between the point clouds according to the coordinates and curvature, and finally performs the clustering segmentation according to the comparison relationship between this distance and the set search radius. The experimental results show that this method achieves the fine segmentation of point cloud data at the parts with complex geometric features, and can be applied to point cloud data with different degrees of variation in geometric features by adjusting the curvature constraint parameters, which is of positive significance for target identification, feature description and surface reconstruction.