Lossless progressive compression of LiDAR data using hierarchical grid level distribution

Abstract

This letter considers a new approach for the lossless progressive compression of light detection and ranging (LiDAR) data stored within a LAS file (public file format for the interchange of three-dimensional point cloud data), which is used for storing the results of LiDAR scanning. The presented method builds a hierarchical data model for arranging LAS points into different levels in one pass. The higher levels are compressed using variable length and arithmetic coding, whilst the lower levels apply a prediction model of the non-progressive compression schema. The order of the points, as captured by the LiDAR scanner, has to be preserved within each level as better compression ratios are achieved in this way.

Compression and Recovery of 3D Broad-Leaved Tree Point Clouds Based on Compressed Sensing

“…It created a color palette according to the spatial redundancy among color attribute data, and applied K-means clustering method to remove redundancy among adjacent color data. However, existing compression algorithms [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32] of point clouds have several weaknesses: (1) low computational efficiency; (2) high time cost; (3) inability to handle complex point clouds, and (4) the need for full sampling.…”