18 February 2010 Structured light-based high-accuracy depth imaging applied for DIBR in multiview 3DTV
Author Affiliations +
In this paper we address the estimation of scene range for multiview 3DTV applications. Based on their similar setup requirements, we propose a depth camera system combining the strengths of Structured Light and Active Stereo techniques to obtain a fast depth map recovery for a larger variety of scenes and from a wider view with less occlusion. The proposed system is composed by a stereo rig and a projector. A stripe-based one shot pattern is projected into the target scene. Using an efficient decoding technique, reliable correspondences can be found between each camera and the projected pattern. Decoded areas common to both cameras are triangulated and used to calibrate the projector on the fly in order to estimate depth in decoded areas only visible in one camera. Correspondences in undecoded areas are estimated via a stereo matching procedure. The different estimated data is then combined in a single depth map. We introduce a 3 projection method that improves the structured light identification in strong illumination conditions and more robust to color objects surfaces of the scene. We demonstrate the efficacy of the integration method experimental results with special emphasis on its performance in the context of the development and contents creation for depth image-based representation (DIBR) 3DTV.
© (2010) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Roger Blanco Ribera, Taeone Kim, Jinwoong Kim, Wookjoong Kim, Namho Hur, "Structured light-based high-accuracy depth imaging applied for DIBR in multiview 3DTV", Proc. SPIE 7524, Stereoscopic Displays and Applications XXI, 75240J (18 February 2010); doi: 10.1117/12.838756; https://doi.org/10.1117/12.838756


Back to Top