Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Reduced-complexity entropy coding of transform coefficient levels using truncated golomb-rice codes in video compression

: Nguyen, T.; Marpe, D.; Schwarz, H.; Wiegand, T.


IEEE Signal Processing Society:
18th IEEE International Conference on Image Processing, ICIP 2011. Vol.2 : Brussels, Belgium, 11 - 14 September 2011
Piscataway: IEEE, 2011
ISBN: 978-1-4577-1304-0 (print)
ISBN: 978-1-4577-1302-6
International Conference on Image Processing (ICIP) <18, 2011, Brussels>
Fraunhofer HHI ()

In hybrid video coding, the difference between the intra or inter prediction signal and the original signal is transmitted using block-based transform coding. The state-of-the-art in coding the quantized transform coefficients is the approach specified in H.264/AVC for context-adaptive binary arithmetic coding. It has, however, been shown that the number of binary symbols that have to be arithmetically coded for the transform coefficients can become very large, making the concept less attractive for high rate applications. To overcome this issue, we propose a combination of simple variable-length codes and context-adaptive binary coding, which yields the same coding efficiency as the H.264/AVC transform coefficient coding at a lower complexity level and which has been adopted into the HEVC test model (HM).