Compression of Probabilistic Volumetric Models using multi-resolution scene flow

作者：

Highlights：

• A novel algorithm for scene flow estimation over Probabilistic Volumetric Models

• Scene decomposition into coherent motion partitions based on the EM algorithm

• Motion partition re-parametrization as locally affine transformations

• Compression system based on flow, motion partitioning and affine fitting

摘要

•A novel algorithm for scene flow estimation over Probabilistic Volumetric Models•Scene decomposition into coherent motion partitions based on the EM algorithm•Motion partition re-parametrization as locally affine transformations•Compression system based on flow, motion partitioning and affine fitting

论文关键词：Optical flow,PVM,Compression,3-d,Segmentation,Coarse-to-fine,Scene flow,Horn and Schunck,GPU

论文评审过程：Received 18 January 2016, Revised 1 March 2017, Accepted 24 June 2017, Available online 6 July 2017, Version of Record 18 July 2017.

论文官网地址：https://doi.org/10.1016/j.imavis.2017.06.005