摘要:
Provided are a method and apparatus for encoding and decoding a datastream into which multiview image information is inserted. The method of decoding a multiview image datastream includes extracting multiview image information including information on at least one view image of a multiview image, from at least one elementary stream of the multiview image datastream; extracting a multiview image parameter regarding the multiview image based on the number of elementary streams and a correlation between view images of the multiview image; and restoring the multiview image by using the extracted multiview image parameter and the extracted multiview image information.
摘要:
Provided are a method and apparatus for encoding and decoding a stereoscopic image. A stereoscopic image restoring method includes parsing a received data stream into image data of a stereoscopic image and information regarding the stereoscopic image; extracting a camera parameter depending on individual characteristics of each of one or more cameras which have captured the stereoscopic image; and decoding and restoring the image data of the stereoscopic image.
摘要:
A method and apparatus for reproducing a stereoscopic image are provided. The method includes generating information about parallax between a left-eye image and a right-eye image of objects included in the stereoscopic image based on a stereo camera parameter of the stereoscopic image, controlling the depth of the objects in the stereoscopic image based on the generated information, and reproducing the stereoscopic image based on the controlled depth. Accordingly, the stereoscopic image can be reproduced while minimizing the eye fatigue of a viewer.
摘要:
Provided are a method and apparatus for receiving and generating an image data stream including a three dimensional (3D) image. The method of receiving an image data stream includes receiving an image data stream including at least one of two dimensional (2D) and 3D image data periods; extracting local 3D image parameters, which are parameters of each image data period, from the image data stream; and restoring at least one of 2D and 3D images by using the local 3D image parameters. In the method, each 3D image is composed of at least one of a base image and an additional image, and the local 3D image parameters include stereoscopic arrangement order information representing an arrangement order of the base image and additional image of the 3D image.
摘要:
Provided is a method of decoding a stereoscopic image data stream. In the method, a plurality of monoscopic image data regions are determined from among image data regions of a stereoscopic image data stream; a composition time stamp (CTS) of each of the monoscopic image data regions is extracted; and, if there are colliding image data regions having the same CTS from among the monoscopic image data regions, monoscopic image data is extracted from a monoscopic image data region of a main stream of the colliding image data regions.
摘要:
Provided is a method and apparatus for determining stereo views of a stereoscopic image. The method includes: selecting a pair of first and second images from a pair of frames of the stereoscopic image, the pair of frames including a base view image and an additional view image. The method further includes determining which of the first image and the second image is the base view image and which of the first image and the second image is the additional view image based on a parallax between the pair of first and second images.
摘要:
Provided is a method of receiving multiview camera parameters for a stereoscopic image. The method includes: extracting multiview camera parameter information for a predetermined data section from a received stereoscopic image data stream; extracting matrix information including at least one of translation matrix information and rotation matrix information for the predetermined data section from the multiview camera parameter information; and restoring coordinate systems of multiview cameras by using the extracted matrix information.
摘要:
Provided are a method and apparatus for generating a stereoscopic image bitstream. The method includes dividing a base view image and an additional view image into blocks having a predetermined size, generating a combined image by combining the blocks of the base view image and the blocks of the additional view image, recording the combined image in a payload area of the stereoscopic image bitstream, and recording block combination pattern information about a pattern of arranging the blocks of the base view image and the additional view image in the combined image, in a header area of the stereoscopic image bitstream.
摘要:
A method of transforming a 3D video format of a 3D video, the method including receiving a video sequence comprising 3D video that includes left-viewpoint video and right-viewpoint video; estimating at least one of disparity information between the left-viewpoint video and the right-viewpoint video and correlation information between neighboring pixel values of the left-viewpoint video and the right-viewpoint video, and determining a 3D video format of the 3D video based on a result of the estimating; transforming the left-viewpoint video and the right-viewpoint video into a format, based on the determined 3D video format; and displaying the transformed left-viewpoint video and the transformed right-viewpoint video three-dimensionally on a the display device.
摘要:
A method of determining a two-dimensional (2D) or three-dimensional (3D) display mode is provided. An image sequence is received. Whether a current image included in the image sequence is a 2D or 3D image is determined. Based on a result of the determination, a 2D or 3D display mode for the image sequence is determined.