电子学报2018，Vol.46Issue(2)：447-455,9.DOI:10.3969/j.issn.0372-2112.2018.02.026
具有容错能力的L1最优化半自动2D转3D
Error-Tolerant Semi-Automatic 2D-to-3D Conversion via L1Optimization
摘要
Abstract
Sparse-to-dense depth conversion is an important task in semi-automatic 2D-to-3D conversion. Existing methods do not handle structural difference between texture image and depth map,and the error-tolerance of 2D-to-3D is not considered.Inspired by compressive sensing studies,we address these problems in an optimization framework via L1norm. First,data term is built with L1norm to measure the fidelity between estimated depth and user assigned depth.Second,local regularized term is defined by using feature weighted L1norm to measure difference between local neighboring pixels.Third, super-pixels are generated from input image and global regularized term is introduced by using feature weighted L1norm to measure difference between representative pixels from these super-pixels.Then,the energy function for sparse-to-dense depth conversion is defined based on the data term,local regularized term and global regularized term.The split Bregman algorithm is used to solve the energy.Experimental comparisons with optimization based interpolation,random-walks,hybrid graph-cuts and random-walks,soft segmentation constrained interpolation and nonlocal random-walks show that our method dem-onstrates significant advantages over hole and ghosting artifacts for viewpoint synthesis.The PSNR is improved by more than 0.9 dB compared with these methods when user assigns error depth.关键词
2D转3D/最优化/随机游走/图割/L1范数Key words
2D-to-3D conversion/optimization/random-walks/graph-cuts/L1norm分类
信息技术与安全科学引用本文复制引用
袁红星,安鹏,吴少群,郑悠..具有容错能力的L1最优化半自动2D转3D[J].电子学报,2018,46(2):447-455,9.基金项目
国家自然科学基金(No.61671260,No.61502256) （No.61671260,No.61502256）
浙江省自然科学基金(No.LY16F010014,No.LY15F020011,No.LQ14F010001) （No.LY16F010014,No.LY15F020011,No.LQ14F010001）
浙江省教育厅科研项目(No.Y201533511) （No.Y201533511）
宁波市自然科学基金(No.2017A610109,No.2013A610114) （No.2017A610109,No.2013A610114）
