Graphics Reference
In-Depth Information
3.5 Performance analysis of different encoding
approaches for colour plus depth based 3D video and
comparison of left and right view encoding vs. colour plus
depthmapvideoencoding
The following encoding configurations are used in this evaluation;
• Layered encoding with H.264/AVC Scalable Video Coding
• MPEG-4 MAC Encoding
• H.264/AVC Encoding with side-by-side stereoscopic 3D video (not frame com-
patible format)
The Orbi and Interview test sequences are utilised to obtain the Rate-Distortion (R-
D) results. The experiments are carried out using CIF (Common Intermediate Format,
352×288)formatimagesequences in order to evaluate the performance of low bitrate 3D
video applications. As the original resolution of these sequences is 720×576, the image
sequencesareinitiallycropped(i.e.704×576)andthendown-sampledintoCIF(352×288)
format video. During the down-sampling process, four nearest pixel values of the original
format are averaged to represent the pixel value of the down-sampled image.
