Kinect depth map restoration using modified exemplar-based inpainting

Depth maps generated by Kinect cameras often contain a significant amount of missing pixels and strong noise, limiting their usability in many computer vision applications. In this paper we present an efficient hole filling and damaged region restoration method that improves the quality of the depth maps obtained with the Microsoft Kinect device. The proposed approach based on a modified exemplar-based inpainting and adaptive median filtering by exploiting the correlation between color and depth values in local image neighborhoods. Several examples considered in this paper show the effectiveness of the proposed approach for large holes removal as well as recovery of small regions on several test images of depth maps.