Interfered depth map recovery with texture guidance for multiple structured light depth cameras
作者:
Highlights:
• We propose a depth recovery scheme for multiple SLDC systems.
• Interference causes depth loss rather than incorrect depth values.
• Different methods are applied to flat and boundary interfered regions.
• Our method improves PSNR of synthetic data by 4–14 dB.
• Our method keeps flat regions smooth and preserve sharp depth discontinuities well.
摘要
Highlights•We propose a depth recovery scheme for multiple SLDC systems.•Interference causes depth loss rather than incorrect depth values.•Different methods are applied to flat and boundary interfered regions.•Our method improves PSNR of synthetic data by 4–14 dB.•Our method keeps flat regions smooth and preserve sharp depth discontinuities well.
论文关键词:Structured light depth camera,Depth recovery,Markov random field,Discrete Poisson equation,3D video
论文评审过程:Received 9 June 2014, Revised 12 October 2014, Accepted 12 November 2014, Available online 5 December 2014.
论文官网地址:https://doi.org/10.1016/j.image.2014.11.004