A progressive point cloud simplification algorithm with preserved sharp edge data

Due to recent advances in high-speed 3D laser-scanning technologies, the set of dense points collected from the external boundary surface of a physical object, often referred to as the point cloud data, is emerging as a new representation format of 3D shapes. A typical point cloud data set contains millions of coordinate data points, and this leads to significant computational challenges for the subsequent data processing tasks in practical applications. This paper presents a new point cloud simplification algorithm to reduce the number of data points scanned from a mechanical part, in which the boundary surfaces often contain sharp edges. Because of the distinct feature represented by data points located on or near the sharp edges (edge points), these points should always be retained by the simplification process. The proposed algorithm thus identifies these edge points first and then progressively removes the least important data point until the specified data reduction ratio is reached. The quantification of a point’s importance is based on points in its neighborhood and corresponds to the point’s contribution to the representation of local surface geometry. The effectiveness of the proposed algorithm is demonstrated through the simplification results of several practical point cloud data sets.