Videos MeiChen Yeh Outline Video representation Basic video

Videos Mei-Chen Yeh

Outline • Video representation • Basic video compression concepts – Motion estimation and compensation Some slides are modified from Prof. Chu’s slides.

Videos • A natural video stream is continuous in both spatial and temporal domains. • A digital video stream sample pixels in both domains.

Video processing YCb. Cr

Video signal representation (1) • Composite color signal – R, G, B – Y, Cb, Cr • Why Y, Cb, Cr? – Backward compatibility (backand-white to color TV) – The eye is less sensitive to changes of Cb and Cr components Luminance (Y) Chrominance (Cb + Cr)

Video signal representation (2) • Y is the luma component and Cb and Cr are the blue and red chroma components. Y Cb Cr

Sampling formats (1) 4: 4: 4 4: 2: 2 (DVB) 4: 1: 1 (DV) Slide from Dr. Ding

Sampling formats (2) 4: 2: 0 (VCD, DVD)

TV encoding system (1) • PAL – Phase Alternating Line, is a color encoding system used in broadcast television systems in large parts of the world. • SECAM – (French: Séquential Couleur Avec Mémoire), is an analog color television system first used in France. • NTSC – National Television System Committee, is the analog television system used in most of North America, South America, Burma, South Korea, Taiwan, Japan, Philippines, and some Pacific island nations and territories.

TV encoding system (2)

Other frame formats • CIF (352 x 288) • QCIF (176 x 144) Common Interchange Format Quarter Common Interchange Format

Uncompressed bitrate of videos Slide from Dr. Chang

Outline • Video representation • Basic video compression concepts – Motion estimation and compensation

Video compression • A video is a sequence of correlated frames. • Compression: remove redundancy – Make use of the temporal correlation! – The previous reconstructed frame is used to generate a prediction for the current frame.

Video compression • Intra coding: Compress one frame without referring to other frames ~ Image compr e – Each pixel has to be compressed – DCT → Quantization → Entropy coding ssion • Inter coding: Compress one frame depending on the previously encoded frame – Only compress moving objects and new objects – Background can be ignored

Example 1 2 3 1. Compress 2. Compress the motion of and 4 in frame 1. in remaining frames. Direction and magnitude Slide from VC Lab

Motion compensation (1) • Objects tend to move between frames Slide from Prof. Bernd Girod

Motion compensation (2) • Block-based motion compensation – Divide a frame into blocks – Search the previous reconstructed frame for the most matching block – Motion vector: the offsets between block locations search area

Motion compensation: Example Slide from Prof. Bernd Girod

Motion-compensated decoder Residual MV Coded Bitstream Reconstructed frame IDCT VLD Motion Compensation Previous Frame memory Slide from VC Lab

Motion-compensated encoder ②Residuals ③ DCT + Q i. DCT + i. Q Motion Compensated Predictor ④ Previous Frame Buffer ① MV = (dx, dy) Slide from VC Lab

Three types of coded picture • I frame (Intraframe) – Encoded without any temporal prediction • P frame (Forward Predicted Frame) – Encoded using motion prediction from the previous I or P frame • B frame (Bidirectionally Predicted Frame) – Encoded using interpolated motion prediction between the previous I or P frames and the next I or P frames

Example: A typical Group of Picture (GOP) in MPEG-2

Summary • Video compression concepts – Cb, Cr sampling – Motion compensation and estimation • The encoder mimics the behavior of the decoder • Unbalanced computing: searching for a good prediction can be time consuming – GOP structure