Kinect Depth Inpainting via Graph Laplacian with TV21 Regularization

Depth maps provided by Microsoft Kinect often contain large dark holes around depth boundaries and occasional missing pixels in non-occluded regions, as well as noise, which prevent their further usage in real-world applications. In this paper, we present a graph Laplacian based framework to restore missing pixels based on the strong correlation between color image and depth map. To preserve sharp edges and remove noise, the TV₂₁ (Total Variation) prior of depth maps is then integrated as an additional regularizer to the framework. Finally, an efficient and effective iterative optimization method with a closed-form solution at each iteration is presented to address this issue. Experiments conducted on both real scene images and synthetic images demonstrate that our approach gives better performance than commonly-used depth in painting schemes.