关键词: 压缩感知, 点云精简, 分块稀疏, 匹配追踪

Abstract:

With higher precision of the laser 3D scanning, the growing point cloud model data lead to the difficulty of data storage and transmission, thus limiting its application in the body network. In order to solve this problem, the compressed sensing technique is applied to the compression and reconstruction of human body point clouds model to perform data simplification processing by the improved three-dimensional grid method before compression. In light of the characteristics of human point cloud data, the block sparse transformation and orthogonal matching pursuit algorithm is applied to reconstruct the original model. The final experimental error is only  , which confirms the validity and feasibility of the algorithm.