Fast H.264 to HEVC transcoder based on post-order traversal of quadtree structure

Abstract

In this paper, we propose a fast mode decision framework and a fast motion estimation algorithm for H.264 to High Efficiency Video Coding (HEVC) transcoding. The fast mode decision framework employs a post-order (bottom-up) traversal of the coding tree unit (CTU) quadtree. Based on this traversal and H.264 information, several strategies are proposed to reduce HEVC modes to be tested and a ratedistortion (RD) cost prediction model is used to terminate the processing of a tested mode early. The proposed fast motion estimation algorithm selects the best candidate from a list of H.264 motion vectors (MVs) and previously encoded HEVC MVs. Compared to a full re-encoding, experimental results show that the proposed solution achieves speed-ups of up to 12.75x, for an average BD-Rate of 3.28%.

Efficient H.264-to-HEVC Transcoding Based on Motion Propagation and Post-Order Traversal of Coding Tree Units

“…x,y,k represent the same differences for the corresponding 2 × 2 chroma blocks. Contrary to the original HM and our previous transcoder [15], the chroma blocks are considered in this equation (in addition to the luma block) to improve the prediction accuracy.…”

“…In contrast, the fast transcoder based on content modeling and early termination (CM+ET) proposed by [5] Peixoto et al achieve an average speed-up of 3.83x and an average BD-Rate of 7.56% [5] for similar test conditions. Table III presents a comparison between our MP+FMD approach, our previous approach [15] and the CM-ET approach [5] for four sequences. Results show our MP+FMD approach increases the speed-up with a low impact on the coding efficiency compared to our previous work.…”

“…In our previous work [15], we proposed an H.264-to-HEVC transcoder combining a fast motion propagation algorithm and a fast mode decision framework. The motion propagation algorithm creates an MV candidate list at the CTU level.…”

“…Comparisons of the proposed transcoder, our previous work[15], and the CM+ET approach[5], for a low delay P setting with one reference on 416×240 (Class B) sequences…”

Inter-Prediction Optimizations for Video Coding Using Adaptive Coding Unit Visiting Order

“…Notice that a bottom-to-top approach has already been used before in video coding, but in the context of intra-prediction, to visit the transform blocks in the recursive structure used for transform and quantization [28], or to calculate gradient intensity of the CU [29]. Recently, it has also been used in the context of transcoding from AVC to the HEVC [30].…”