Study Of Advanced Transforms In HEVC/H.265

HEVC/H.265 is the newest video coding standard currently.The growing popularity of high-definition video brings challenges and pressures to traditional video coding standards,and HEVC is proposed to deal with this situation.The ultimate goal of the HEVC standard is to reduce by half the bit-rate for the equal video frame quality with the H.264 / AVC standard.In order to achieve higher compression efficiency,a plurality of small advanced techniques have been utilized in the coding elements of HEVC design.This paper focuses on the improvement and optimization of the transform algorithm in the intrapicture coding of HEVC.High efficiency video coding(HEVC)defines 35 prediction modes in its intra prediction stage to signal the direction information of residual blocks.Traditionally,separable two-dimension(2-D)transforms(integer DCT and DST)are utilized in a similar manner as in the previous H.264/AVC standards.2-D transforms used in HEVC are computed by applying two 1-D transforms in the horizontal and vertical directions.However,such 2-D transforms cannot yield the best energy compaction for a 2-D directional source where the dominating directional information is other than the horizontal or vertical one.In this case,neither DCT nor DST can cover the varieties of video signals.In order to overcome this drawback,we build an elliptical model with directionality and design some nonseparable transforms based on the Karhunen-Lo`eve transform in this paper.Specifically,we derive a non-separable transform in closed-form for each intraprediction mode and replace the default transform in HEVC.Simulation results reveal that compared to the original HEVC coding in HM-16.4,1.7% and 2.0% on average and up to 7.7% and 8.4% BD-rate reduction can be achieved for luma and chroma component,respectively.In the meantime,the test results show that both the encoding time and decoding time increase only about 5%.