Interactive Multiview Video Communication System

2014 Fiscal Year Final Research Report

• Project/Area Number: 23700136
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Media informatics/Database
• Research Institution: National Institute of Informatics
• Principal Investigator: CHEUNG GENE 国立情報学研究所, コンテンツ科学研究系, 准教授 (40577467)
• Project Period (FY): 2011-04-28 – 2015-03-31
• Keywords: 画像処理 / グラフ信号処理

Outline of Final Research Achievements

Today's imaging technologies can now capture texture maps (e.g., color images) and depth maps (per-pixel distance between objects in the 3D scene and camera) from multiple viewpoints simultaneously. Given this rich set of input images, the objective is to develop key technologies that enable a range of interactive 3D imaging applications. In particular, we focus on three aspects of an end-to-end visual communication system: i) coding of 3D images, ii) transmission of 3D data over data networks, and iii) restoration of virtual views given transmitted 3D data. For coding of 3D data, we have developed is a multi-resolution edge-adaptive image codec based on graph Fourier transform (GFT). For transmission of 3D data, we have developed a strategy to exploit the data redundancy inherent in the representation of multiview texture-plus-depth format for loss recovery. For restoration of 3D images, we have studied the use of graph Laplacian regularizer as signal prior for regularization.

Free Research Field

信号処理