Inter-Frame Compression
of 3D Point Cloud Sequences

Ji-Su Kim
jisukim@mcl.korea.ac.kr
Korea University

Seonho Lee
seonholee@mcl.korea.ac.kr
Korea University

Jae-Han Lee
jaehanlee@mcl.korea.ac.kr
Korea University

Chang-Su Kim
changsukim@korea.ac.kr
Korea University

Abstract

In this paper, we propose a 3D point cloud sequence compression algorithm based on inter-frame coding. First, we create an octree structure for the 3D point cloud. For the first frame, we compress geometric and color attributes through intra-frame coding. Then we apply inter-frame coding to the rest of the frames. We compress the geometric attributes using a dual octree and compress the color attributes using the k-means clustering. Both intra-frame coding and inter frame coding are performed with arithmetic codes. Experimental results demonstrate that the proposed algorithm compresses point cloud sequences more efficiently than the algorithm using only intra-frame coding.

Results

Publications

Ji-Su Kim, Jae-Han Lee, and Chang-Su Kim, "3D Point Cloud Sequence Compression Using Block Skipping," in Proc. International Workshop on Frontiers of Computer Vision (IWFCV), Gangneung, Korea, Feb. 2019. [pdf] [bibtex]

Ji-Su Kim, Seonho Lee, Jae-Han Lee, and Chang-Su Kim, "Inter-Frame Compression of 3D Point Cloud Sequences," Proc. International Conference on Electronics, Information, and Communication (ICEIC), Auckland, New Zealand, Jan. 2019. [pdf] [source code(coming soon)] [bibtex]