A switching morphological algorithm for depth map recovery

Abstract

In this paper, we propose a switching morphological filter for RGB-D depth map recovery. The switching algorithm consists of the following steps: detection of noisy pixels and hollow areas (holes) using morphological filtering; correction of the detected noisy and hole pixels. With the help of computer simulation, we show that the proposed algorithm is able to fast recover depth maps. So, the accuracy of 3D surface reconstruction with the proposed filtering noticeably increases. The performance of the proposed algorithm is compared in terms of the accuracy of 3D surface reconstruction and processing time with that of common depth filtering algorithms.