2D Reprojection of Plenoptic 3D Voxel Data Using Gaussian Intensity Spreading

Seong-Jun Bae, Soowoong Kim, Hahyun Lee, Jinho Lee, Sung-Chang Lim, Gun Bang, Jung Won Kang

Abstract: This paper presents an efficient way of 2D reprojection method of 3D voxel data which is generated from multiple 2D image sources. In order to minimize the image quality degradation generated by postprocessing such as hole filling, A 2D reprojection method using gaussian aggregation of voxel attribute intensity is proposed. Proposed method efficiently removes generation of holes while maintaining graceful intensity of projected 2D image planes, which can finally reconstruct original 2D source images with less quality degradation.