Real-Time Depth Image Acquisition and Restoration for Image Based Rendering and Processing Systems

Depth information is an important ingredient in image-based rendering (IBR) systems. Traditional depth acquisition is mainly based on computer vision or depth sensing devices. With the advent of electronics, low-cost and high-speed depth acquisition devices, such as the recently launched Microsoft K...

Full description

Saved in:
Bibliographic Details
Published in:Journal of signal processing systems 2015-04, Vol.79 (1), p.1-18
Main Authors: Wang, Chong, Zhu, Zhen-Yu, Chan, Shing-Chow, Shum, Heung-Yeung
Format: Article
Language:English
Subjects:
Acquisitions
Circuits and Systems
Computer Imaging
Devices
Electrical Engineering
Engineering
Filtering
Image acquisition
Image Processing and Computer Vision
Marketing
Mathematical models
Pattern Recognition
Pattern Recognition and Graphics
Rendering
Restoration
Signal processing
Signal,Image and Speech Processing
Vision
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Depth information is an important ingredient in image-based rendering (IBR) systems. Traditional depth acquisition is mainly based on computer vision or depth sensing devices. With the advent of electronics, low-cost and high-speed depth acquisition devices, such as the recently launched Microsoft Kinect, are getting increasingly popular. A comprehensive review of these important and emerging problems and their solutions are thus highly desirable. This paper aims to 1) review and summarize the various approaches to depth acquisition and highlight their advantages and disadvantages, 2) review problems arising from calibration and imperfections of these devices and state-of-the-art solutions, and 3) propose a surface-normal-based joint-bilateral filtering method for fast spatial-only restoration of missing depth data and a confidence-based IBR algorithm for reducing artifacts under depth uncertainties. For the latter, we propose a confidence measure based on color-depth, spatial and restoration information. A joint color-depth Bayesian matting approach is proposed for refining the depth discontinuities and the alpha matte for rendering. Improved rendering results are obtained compared with rendering using conventional restored depth maps. Possible future work and research directions are also briefly outlined.
ISSN:1939-8018
1939-8115
DOI:10.1007/s11265-013-0819-2