Electronics and Telecommunications Research Institute

Real Image Corp.

Venture 3D, Inc.

Jong-Ryul Kim

Nam-Ho Hur

Yung-Lyul Lee

Soo-In Lee

Suk-Hee Cho

Dae-Hee Kim

Hyoung-Woo Kim

Kwan-Woo Kim

Ju-Hyae Kim

Provided are an apparatus and method for predictive coding/decoding for improving a compression rate of multiview video using one or two additional reference frame buffers. The predictive encoding apparatus includes: a multiview reference picture providing unit for providing a reference picture for a predictive encoding according to temporal and spatial GOP structure information; a prediction unit for creating a vector by predicting which part of the reference picture inputted from the multiview reference picture providing unit is referred by a picture to currently encode; a transforming and quantizing unit for obtaining a difference signal between the predicted signal inputted from the prediction unit and a picture signal to currently encode, transforming the obtained difference signal, quantizing the transformed signal, and compressing the quantized signal; and an entropy encoding unit for encoding the quantized signal and the vectors according to a predetermined scheme and outputting the encoded signal.

Apparatus of predictive coding/decoding using view-temporal reference picture buffers and method using the same

Provided are a method and apparatus for generating a stereoscopic image from a two-dimensional (2D) image by using a mesh map and a computer readable recording medium having recorded thereon a computer program for executing the method. Also provided are a method and apparatus for generating a stereoscopic image by reading a 2D image, displaying the 2D image and a mesh map by overlapping the 2D image and the mesh map, and editing mesh shapes and depth information (depth values) of the mesh map by a user, and a computer readable recording medium having recorded thereon a computer program for executing the method. The method of generating a stereoscopic image includes receiving a 2D image; displaying the 2D image and a mesh map by overlapping the 2D image and the mesh map; editing mesh shapes and depth information (depth values) of the mesh map by a user in accordance with shapes of a displayed image; calculating relative depth information of pixels included in the 2D image in accordance with the mesh shapes and the depth information of the edited mesh map; and generating a stereoscopic image file by using the calculated relative depth information of the 2D image. The present invention may be used in a system for generating a stereoscopic image from a 2D image including a general still image or moving picture.

Method and apparatus for generating stereoscopic image from two-dimensional image by using mesh map

Growing community of inventors

Seoul, South Korea

Jong-Ryul Kim