| Autor: |
I. Manju, K. S. Srinivasan, R. Haresh, E. Rohith Kumar |
| Rok vydání: |
2020 |
| Předmět: |
|
| Zdroj: |
Proceedings of International Conference on Recent Trends in Machine Learning, IoT, Smart Cities and Applications ISBN: 9789811572333 |
| DOI: |
10.1007/978-981-15-7234-0_77 |
| Popis: |
Sample adaptive offset (SAO) adapted in high efficiency video coding (HEVC) standard is an in-loop filtering technique to reduce mean sample distortion occurring due to artifacts introduced by using larger transforms. The SAO involves classifying each sample into edge and band, calculates and adds appropriate offset values to the samples. In VLSI implementation of SAO, band offset (BO) calculation consumes a comparatively larger area and gate count. Some previously proposed algorithm for reducing the total bands from 32 to lower number of bands to reduce BO area compromises the proper prediction and selection efficiency of the four consecutive band group for which offset is needed to be calculated and signaled. This paper proposes a preselection technique using which from the total number of 32 bands, only 16 bands contributing majority samples in given coding tree unit (CTU) are selected and passed on for further processing. The usage of 16 bands provides sufficient space for making better prediction and selection of the required four consecutive band group along with optimum reduction in the consumed area. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
