Analysis of Transform unit of HEVC and Earlytermination of Partitions

Abstract

The High efficiency video coding (HEVC) is the latest video coding standard developed by the Joint Collaborative Team on Video Coding (JCT-VC). The compression performance of HEVC is higher than AVC and other previous standards. In HEVC video compression is 50% more efficient than AVC having same video quality at half bitrate. The quad-tree partition of the coding unit (CU) in HEVC accounts for a significant amount of the encoding difficulty. In the proposed paper we have presented RasterScan approach for pruning of Quad-Tree structure of the transform unit (TU) to reduce the computational burden of HEVC encoder at intra-prediction mode. The CU is partition into Quad- Tree structure based on Rate Distortion cost (RD) computation. The CU partitioning structure is scanned by using Raster Scan approach. During scanning the depth of CU partitioning is extracted by comparing each CU partition pattern with its predecessor. The extracted depth level is considered as the best depth level for the TU. once the best TU depth is identified further partitioning of the TU within the given CU is stopped. pruning of Quad-Tree of TU reduces the computational burden of HEVC and measured in terms of reduction in the encoding time. Encoding time of proposed model is reduced by 37.62% as compared to HEVC Test Model (HM) at negligible loss of PSNR and Bitrate.