Context Adaptive Variable Length Coding
Context Adaptive Variable Length Coding ( CAVLC , German: Context Adaptive Code of Variable Length) describes a context-dependent run length coding and is a type of lossless entropy coding . CAVLC is used for video compression MPEG-4 / Part 10 ( H.264 / AVC). The CAVLC was designed in such a way that it can use as many peculiarities of MPEG quantization as possible. Features of MPEG quantization include:
- The initial set mainly contains zeros,
- the highest values are at the beginning of the initial quantity and become lower and at the end
- the highest values in the initial set are often sequences of 1 and −1.
A data stream compressed with CAVLC is around 10-20% larger than a data stream compressed with the more computationally intensive Context-Adaptive Binary Arithmetic Coding (CABAC).
literature
- Iain EG Richardson: H.264 and MPEG-4 Video Compression: Video Coding for Next-generation Multimedia . John Wiley & Sons Ltd., Chichester 2003.
Web links
- Iain EG Richardson, Variable Length Coding H.264 / MPEG-4 Part 10 White Paper , English
- Jie Liang, Simon Fraser University, CAVLC in H.264, PDF lecture on CAVLC , English
- JVT Document JVT-C028, G. Bjontegaard and K. Lillevold, “Context-Adaptive VLC Coding of Coefficients”, Fairfax, VA, May 2002
- Differences in coding efficiency between CABAC and CAVLC, shown using an example image