Projection based dynamic point cloud compression using 3DTK toolkit and H.265/HEVC

This paper presents novel compression method for dynamic point clouds based on projections and the H.265/HEVC video coder. We used 3DTK - The 3D Toolkit to create equirect-angular projection images and x265 as H.265/HEVC coder, to compress and extract created projection images. Compression was introduced in 3DTK generated projection image size (different pixel size, but compatible with later video compression), as well as lossless/lossy video compression. Visual inspection shows better results for compression only using different projection resolution, with lossless video compression. Lossy video compression adds noise and creates additional points, resulting in lower visual quality.