Video Compression (Spatial & Temporal)

💡 The Core Idea

Raw video has a massive bitrate. Compression reduces size by exploiting Spatial Redundancy (pixels within a frame are similar) and Temporal Redundancy (adjacent frames are similar).

🧠 Mechanisms

Spatial (Image) Compression (JPEG)

1. Transformation: Convert RGB to YCbCr (Luminance/Chrominance).
2. DCT: Apply Discrete Cosine Transform to 8x8 blocks to separate high/low frequencies.
3. Quantization: Divide coefficients by a quantization table ($Q$) to reduce precision of high frequencies (Lossy step): $F_{quantized}(u, v) = \frac{F(u, v)}{Q(u, v)}$ .

Temporal Compression (MPEG)

Instead of sending every full image, send differences.

• I-Frame (Intra): A full, self-contained image (JPEG encoded).
• P-Frame (Predicted): Encodes only the changes from the previous frame.
• B-Frame (Bi-directional): Encodes changes relative to both past and future frames.
• GoP: A “Group of Pictures” is the sequence between I-frames.

🔗 Connections

• Source: Source - Multimedia Applications