Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Inter-Component Transform for Color Video Coding

: Rudat, C.; Helmrich, C.R.; Lainema, J.; Nguyen, T.; Schwarz, H.; Marpe, D.; Wiegand, T.


Institute of Electrical and Electronics Engineers -IEEE-; IEEE Circuits and Systems Society:
Picture Coding Symposium, PCS 2019 : Ningbo, China, 12 - 15 November 2019
Piscataway, NJ: IEEE, 2019
ISBN: 978-1-7281-4705-5
ISBN: 978-1-7281-4704-8
Picture Coding Symposium (PCS) <34, 2019, Ningbo/China>
Fraunhofer HHI ()

In natural digital images and videos, correlations between color components can be observed. These correlations can be exploited to achieve additional coding gain in modern block-based hybrid video coding. To this end, we propose the use of a block-wise, rotational inter-component transform (ICT) applied to the two residual chroma signals that result from conventional intra or inter-picture prediction. Different ICT parameterizations in terms of number and quantization of the rotational angles as well as resulting components signaled in the coded bitstream are investigated. An implementation into the currently developed Versatile Video Coding (VVC) reference software provides average bitrate savings of up to 0.7% (All Intra configuration) with negligible increases in implementation complexity and runtime. Our proposal has been adopted into the VVC draft specification text.