Tomoko Sawabe

Beyond 4K: 8K 60p live video streaming to multiple sites

The high definition of 4K motion pictures makes them applicable to a wide variety of purposes. The development of 4K image equipment such as cameras, displays, and playback systems has made easy viewing of 4K video possible. Video frame rates have also become higher, enabling 4K-60 fps progressive (60p) video streams to be transmitted over IP networks. However, problems remain in increasing video image resolution to 8K or more so as to build systems in which many people collaborate at the same time. This paper describes a video transmission system that attains image resolution of 8K-60p or higher by synchronizing multiple 4K transmission systems we have developed. Since 8K-60p transmission over IP networks has extremely high traffic rates, schemes for achieving robust transmission such as forward error correction (FEC) must also have very high rates. We have also developed techniques for attaining reliable transmission, i.e., the use of low-density generator matrix (LDGM) codes to achieve high throughput FEC, application-layer multicasting, and monitoring network status at multiple points.

Digital Cinema and Super-High-Definition Content Distribution on Optical High-Speed Networks

Digital cinema is a promising application that utilizes high-speed optical networks to transfer super-high-definition (SHD) images. The networks are primarily used for distributing digital cinema contents in packet data form, and are also used to support new services such as the live streaming of musicals and sport games to movie theaters. While current transfer services offer high-definition (HD) quality video, live-streaming applications will soon shift to providing cinema quality 4K content to both business and movie theaters users. The extra-high-quality 4K format enables a realistic telepresence, and will be combined with special tools such as video editing systems to realize effective remote collaboration for business workspaces. This paper introduces successive research on SHD image transmission and its application, especially in digital cinema and associated application fields.

Transmission characteristics of MPEG2 encoded super high definition images

We have investigated super high definition (SHD) images supporting resolutions beyond 2048/spl times/2048 pixels/frame with 24-bit color separation at 60 frames/s for many professional applications of multimedia communications. However, 6 Gbps is needed to transmit moving SHD image sequences without any compression. This paper expands the MPEG2 standard from main profile at high level (MP@HL) to main profile at super high level (MP@SHL) to transmit SHD images via B-ISDN networks. The transmission of MPEG2 compressed SHD images over ATM networks is simulated. The strong burstiness of compressed SHD images is clarified and the transmission characteristics of these images over 155 Mbps and 600 Mbps ATM lines are determined.

Implementation of super high definition image processing on HiPIPE

In order to manipulate SHD (super high definition) images, a novel parallel processing unit called HiPIPE (Highly Parallel Image Processing Engine) is developed as a project of NOVI-II. The engine is connected to an SHD image display unit and an image data storage unit. Extremely high computational power is obtained by a multicomputer type parallel processing technique. 128 processing elements are connected by a mesh network. Various image coding schemes are carefully explored from the viewpoint of parallel processing, and the problem of processor connections is examined. A novel load-balancing technique, called 2-dimensional butterfly data shuffling, is developed and implemented. The current version of HiPIPE uses just 128 scalar processing elements and has more than twice the power of a single-processor CRAY-2 for a still SHD image coding task.<<ETX>>

A study on non-octave scalable coding with filter bank and its performance evaluation using EBCOT

JPEG2000, an international standard for still image compression, has various features, such as high coding performance, unified lossless/lossy compression, resolution scalability and SNR scalability. Resolution scalability is an especially promising feature. Unfortunately, the current implementation of resolution scalability is restricted to powers of two. We propose non-octave scalable coding. Two types of non-octave scalable coding are implemented. One is based on a DCT filter bank and the other uses a wavelet transform. The latter model is compatible with JPEG2000, part 2. By using the proposed algorithm, images of various resolutions can be decoded from a compressed bit stream. Experimental results show the effectiveness of the proposed algorithm.

A study on non-octave resolution conversion based on JPEG2000 extensions

The features of JPEG2000, an international standard for still image compression, include (1) high coding performance, (2) unified lossless/lossy compression, (3) resolution and SNR scalability. Resolution scalability is an especially promising feature given the popularity of super high definition (SHD) images like digital-cinema. Unfortunately, its current implementation of resolution scalability is restricted to powers of two. We introduce a non-octave resolution conversion method that is compatible with JPEG2000 part2. By using the proposed algorithm, images of various resolutions can be decoded from a compressed JPEG2000 part2 code stream. Experimental results from digital-cinema test sequences show the effectiveness of the proposed algorithm.

A 12.8 GFLOPS multi-DSP system for super high definition image processing

A multicomputer-type parallel processing system has been developed that offers a peak performance of 12.8 GFLOPS. This system is called NOVI-II HiPIPE, and was designed for super-high-definition (SHD) image processing. SHD image is a new image medium and its resolution is higher than that of HDTV. Because SHD image processing requires computing power that exceeds the ability of any single digital signal processor (DSP), current or anticipated, NOVI-II HiPIPE uses a multicomputer-type parallel processing architecture. A 100-MFLOPS vector processor serves as the backend processor of each processing element. The current version of NOVI-II HiPIPE uses 128 mesh connected processing elements and 128 vector processors. The configuration and performance of NOVI-II HiPIPE are described.<<ETX>>

A study on multiresolution lossless video coding using inter/intra frame adaptive prediction

Lossless video coding is required in the fields of archiving and editing digital cinema or digital broadcasting contents. This paper combines a discrete wavelet transform and adaptive inter/intra-frame prediction in the wavelet transform domain to create multiresolution lossless video coding. The multiresolution structure offered by the wavelet transform facilitates interchange among several video source formats such as Super High Definition (SHD) images, HDTV, SDTV, and mobile applications. Adaptive inter/intra-frame prediction is an extension of JPEG-LS, a state-of-the-art lossless still image compression standard. Based on the image statistics of the wavelet transform domains in successive frames, inter/intra frame adaptive prediction is applied to the appropriate wavelet transform domain. This adaptation offers superior compression performance. This is achieved with low computational cost and no increase in additional information. Experiments on digital cinema test sequences confirm the effectiveness of the proposed algorithm.

Special Issue Image Technology of Next Generation. Design and Performance Evaluation of Fully Programmable Entropy CODEC.

Entropy coding algorithms, such as Huffman coding, frequently use bit operations with variable word lengths and although the pipeline processors of existing DSPs can perform numerical processing efficiently, the bit operations in entropy coding are a heavy burden because of a mismatch between these operations and the pipeline processors. We propose a new approach to implement a full programmable entropy CODEC with a combination of a high-level design LSI-CAD system and field-programmable gate arrays. This method simultaneously provides high performance for entropy coding and full programmability. The effectiveness of the proposed approach is demonstrated by using it in designing the Huffman encoder of a JPEG image coding algorithm.

Design of multiprocessor DSP chip set for superhigh-definition image processing

This paper details the design of a DSP chip set for NOVI-III, a massively parallel digital signal processing system for super high definition (SHD) image coding. In the first step, achieving real-time SHD image processing, we implemented the JPEG coding algorithm on the highly parallel DSP system called NOVI-II HiPIPE, and evaluated its processing performance. The result shows that a real-time CODEC must have an average computational performance of over 100 GFlops (floating point operations) to process SHD motion images. Based on our experience of still SHD image coding, a new DSP chip set is being designed for NOVI-III. The chip set consists of three main chips. The first chip is a vector processor that, in its original form, had a peak performance of 120 MFlops. It is being redesigned to achieve a peak performance of 540 MFlops with 0.5 micrometers fabrication technology. Second is an intercommunication network switch that has six 400 Mbps data links. Third is a RISC type DSP core which controls the vector processor, communication switch and internal memories. This DSP core also has a special function that efficiently performs the bit operations required by the Huffman coding process.