Cost-efficient HEVC-based quadtree splitting (HEQUS) for VVC Video Transcoding

The release of the latest video coding standard, known as Versatile Video Coding (VVC), has created the need to convert current High Efficiency Video Coding (HEVC) content to this new standard. However, the traditional cascade transcoding pipeline is not effective due to the exorbitant computational complexity of VVC. With this in mind, this paper proposes a fast HEVC-VVC transcoder that implements a probabilistic classifier based on Naïve-Bayes at the first partitioning level (128 × 128 pixels). This model uses features extracted from the 128 × 128 size blocks of the residual and reconstructed frames in the HEVC bitstream, and their correlation with the block partitioning structure. For the subsequent VVC coding depth levels, partitioning decisions are derived from the HEVC structure. The results achieve a 57.08% transcoding time reduction with a BD-rate penalty of 2.40%, compared with a traditional transcoding approach for the random access encoding configuration.     

云平台上AVS视频转码系统的设计与实现

随着网络与宽带技术的飞速发展,数字视频呈现出海量化与多样化的特征.AVS作为我国自主音视频标准,编码效率优于同期国际标准,在保证图像质量的同时,便于视频数据的存储与传输.为了将数字视频进行高效的AVS转码,提出并实现了一种云平台上的AVS转码系统,该系统采用音视频分离方法将其他格式视频文件快速转码成AVS格式,并避免了转码文件中音视频内容间不同步问题,实验结果证明了方案的有效性.     

多媒体信息服务系统中的转码技术研究

多媒体信息服务MMS是一项全新的移动数据业务，是一个涉及终端设备制造商、网络运营商、内容和应用提供商的综合性移动解决方案。转码是一种对媒体内容进行调整以满足与特定环境有关的约束与爱好的技术．能使MMS信息在不同的网络与移动终端无缝地传输与表示，是MMS成功应用的关键。文章在介绍转码技术原理的基础上，详细阐述了MMS应用中的三种类型转码的关键技术，并讨论了各种类型转码中要解决的主要问题。     

MPEG-2与MPEG-4间转码的关键技术研究与实现

视频转码技术的出现，解决了视频通信领域数据在不同系统和设备间互通互连的问题。本文主要研究了MPEG-2到MPEG-4的转换编码，提出了运动矢量映射的转码结构，并对其中的关键算法进行了研究和实现。仿真试验表明，本文结构和算法在保证了转换编码灵活性的同时，具有较好的转码效率和鲁棒性。     

基于最佳量化的序列图像编码转换码率控制策略

该文分析了编码转换缓冲区的状态,导出了编码转换缓冲区为防止解码器缓冲区下溢和上溢应满足的条件,建立了序列图像编码转换模型。并根据编码转换缓冲器的状态和信道速率,为待编码帧在图像层上预分配目的序列图像编码比特数,使用DCT系数分布特性来表征图像特性。继而为帧内每一具体宏块选定最佳量化因子,提出了基于最佳量化的码率控制策略,模拟实验表明,该码率控制策略能有效地减少、避免缓冲区出现上、下溢的情况,使输出码率趋于稳定,提高了重建序列图像的信噪比。     

MPEG-4 to H.264 transcoding with frame rate reduction

In this paper, a temporal resolution reduction transcoding method that transforms an MPEG-4 video bitstream into an H.264 video bitstream is proposed. The block modes and motion vectors in the MPEG-4 bitstream are utilized in the H.264 encoder for the block mode conversion and motion vector interpolation methods. Four types of motion vector interpolation methods are proposed in order to avoid the use of brute-force motion estimation in H.264. According to the experimental results, the proposed methods achieve a 3∼4 times improvement in the computational complexity compared to the cascade pixel-domain transcoding method, while the PSNR (peak signal to noise ratio) is degraded by 0.2∼0.9 dB depending on the bitrates.

Metadata-driven multimedia transcoding for distance learning

Advances in multimedia processing capabilities of electronic devices and the rapid growth of the Internet have contributed to the proliferation of collaborative applications such as distance learning (DL) webcasting. A key technical challenge in such DL systems is providing access to rich media content to any user regardless of device capabilities, network heterogeneity, and personal preferences. A novel MPEG-21-based adaptation architecture is presented to overcome these challenges by performing (1) application layer transcoding that adapts the presentation format of DL content to match device media decoding capabilities and user-desired modality, (2) bitstream transcoding to adapt multimedia to match device processing capabilities and encoding bit-rate supported by the network. Experimental results indicate that the proposed system delivers personalized DL content to meet end-user environmental restrictions with small transcoding overhead. This work is partially supported by the Network for Effective Collaboration Technologies through Advanced Research (NECTAR). The authors would like to thank the Knowledge Media Design Institute (KMDI) at the University of Toronto for use of their resources.     

Perfect requantization for video transcoding

Bit rate adaptation is one of the most important types of video transcoding. In this paper, we investigate certain critical points in the spectrum of rate shaping requests. We show that the selection of quantization step sizes may not have monotonic effects on rate-distortion characteristics in the transcoding sense. In other words, rate-distortion tradeoff for transcoding which operates on an already-compressed source can be different than that for regular encoding which operates on an original source. We show in a generic form that careful selections of the step size can lead to much improved performance for transcoding, especially when comparing to what would have been produced through a direct encoding. We demonstrate this unique rate-distortion characteristic through simulations as well as real transcoding scenario with H.264 sequences.

A real-time video watermarking system with buffer sharing for video-on-demand service

Real-time watermarking for streaming video (such as VOD service) requires significant amounts of computing resources. To address this issue, we present a scalable watermarking scheme integrated in a parallel MPEG-2 engine. A content-based block selection algorithm is proposed to efficiently embed the pseudo-random watermark signatures into DCT blocks. Our watermark scheme also provides a robust way to synchronize the watermarked video to the original source at detectors and is very resilient against cumulative and temporal attack.We optimize the parallel watermark engine to achieve real-time watermarking performance. We found that the system throughput could suffer significant degradation when processing high-level MPEG-2 video (such as HDTV) due to inefficient management of memory space. Therefore, we investigated an efficient buffer management scheme consisting of two methods: First we reduced the transmission buffer in slave nodes by frames sharing between frames in the Group-of-Picture (GOP) level. Then we further reduce the buffer space by a dynamic on-demand allocation on the slave side. By solving the memory-shortage bottleneck, the proposed system can support real-time watermarking for multiple high-resolution (up to 1404 × 960) video.     

The Personal Presence System: a wide-area network resource for the real-time composition of multipoint multimedia communications

The Personal Presence System (PPS) is an experimental research prototype platform to support advanced broadband applications, including multimedia conferencing and distance learning. It supports the user's presentation control of advanced network elements (multimedia bridges). The multimedia bridge provides for a single contact point per user, who receives only one combined video stream, which minimizes access bandwidth. This architecture solves the -squared transmission problem of multipoint video conferencing and makes customer terminal equipment simpler and cheaper than high-end desktop workstations. The PPS provides each user with flexible presentation control of a feature-rich video bridge. The PPS also supports advanced features such as object extraction and multimedia object associations.