Search

The High Efficiency Video Coding (HEVC) standardization project mainly aims at improving the compression efficiency beyond the state-of-the-art H.264/AVC standard for high and ultra high definition video contents. The HEVC codec reference software, still under development, encodes video content using a fixed quantization step, thus leading to a variable bitrate stream which may not be suitable for the many multimedia applications where a constant bandwidth is required. Therefore, a rate control algorithm must be integrated in the HEVC codec taking into account that the novel coding tools further reduce the bitrate associated to texture data while increase the bitrate associated to the auxiliary information, e.g. motion vectors and coding modes. In this context, this paper studies the integration of a quadratic modeling rate control algorithm in the HEVC codec. Experimental results reveal that the rate control enabled HEVC codec provides a lower rate-distortion performance regarding a fixed quantization step HEVC codec, notably spending up to 20.12% more bitrate for the same objective quality. However, the integrated rate control algorithm offers lower bitrate fluctuations along the video contents than the HEVC codec with fixed quantization step. [ABSTRACT FROM PUBLISHER]

To face the still growing video compression needs, ITU and MPEG have jointly started a standardization project called High Efficiency Video Coding (HEVC) which aims to improve the compression efficiency of the state-of-the-art H.264/AVC standard for high and ultra high definition video. The HEVC codec still relies on the usual motion compensated, predictive, block based transform coding architecture with block sizes higher than 8×8 being now considered. Furthermore, the codec is also equipped with a Wiener in-loop filter which, together with the H.264/AVC deblocking filter, further reduces the distortion between the original and decoded frames introduced by lossy coding. While this filter allows indeed reducing the sum of square differences between the original and reconstructed frames, it is not specifically designed to reduce the ringing artifacts resulting from the use of large transform block sizes. In this context, this paper proposes to combine an adaptive bilateral filter together with the HEVC Wiener filter. The proposed combined filter allows an average bitrate reduction of about 7% regarding the HEVC codec without the Wiener filter and a 1.5% reduction regarding the HEVC codec with only the Wiener filter, always for the same quality. Moreover, the combined filter also reduces the ringing artifacts according to an objective metric specifically designed to quantify this type of coding artifact. [ABSTRACT FROM PUBLISHER]

The under development High Efficiency Video Coding (HEVC) standard employs several powerful coding tools to obtain improved compression efficiency regarding the state-of-the-art H.264/AVC standard. To efficiently exploit the temporal and spatial redundancies, HEVC adopts a very flexible quadtree coding structure, allowing the encoder to use a block partition better matching the image features. The best combination of HEVC block partitioning and coding modes is found by means of a rate distortion optimization process. During this minimization, the encoder tests all the possible coding modes and block partitions and keeps those providing the smallest RD cost. Due to the large number of available modes and partitions, the minimization involves extremely high complexity, which may not be suitable for real-time application scenarios. To reduce the complexity, two novel fast RDO techniques are proposed in this paper: Top Skip and Early Termination. The experimental evaluation reveals that the proposed techniques significantly reduce the encoding time, notably up to 45% regarding the high complexity HEVC reference software codec, with a negligible quality loss never larger than 0.1 dB. [ABSTRACT FROM PUBLISHER]