We describe a novel approach to filtering coded video data. MPEG video is filtered to reduce frame rate and bandwidth by dropping coded B-pictures. The choice of B-pictures to be dropped is made based on an 'activ...
详细信息
ISBN:
(纸本)0819423564
We describe a novel approach to filtering coded video data. MPEG video is filtered to reduce frame rate and bandwidth by dropping coded B-pictures. The choice of B-pictures to be dropped is made based on an 'activity measure' which indicates the amount of motion within the video scene. We present preliminary results which indicate that this method produces a filtered sequence with higher visual quality than a sequence produced by dropping B-pictures at regular intervals.
This paper describes a coding control scheme for MPEG-2 which maintains the perceived video quality constant, and is suitable for real-time and low latency encoding. We have chosen a proportional integral derivative (...
详细信息
ISBN:
(纸本)0819423564
This paper describes a coding control scheme for MPEG-2 which maintains the perceived video quality constant, and is suitable for real-time and low latency encoding. We have chosen a proportional integral derivative (PID) scheme for the controller, and we adopt the same approach in designing the particular PID feedback function. The main reason for using a PID feedback function is the good tradeoff that it offers between computational complexity, ease of design, and performance. We have chosen to use a new video quality metric called moving picture quality metric (MPQM). This metric models the human visual system and matches subjective evaluations correctly, outperforming existing quality metrics for video. Simulations results are shown for typical video sequences. A comparison with CBR encoding also is presented.
The wavelet transform is defined for infinite-length signals. In practice we only have finite-length signals, so signals must be extended before they can be transformed. The question is how to extend the signal to min...
详细信息
ISBN:
(纸本)0819423564
The wavelet transform is defined for infinite-length signals. In practice we only have finite-length signals, so signals must be extended before they can be transformed. The question is how to extend the signal to minimize signal end effects, or how to find the signal extension that preserves the transform signal length. In this paper we discuss the problem of signal extension in perfect reconstruction filter banks.
The storage and bandwidth requirements of digitalvideo and audio exceed those that can be supported by conventional file servers. Despite the emergence of new compressionalgorithms capable of providing extremely hig...
详细信息
The storage and bandwidth requirements of digitalvideo and audio exceed those that can be supported by conventional file servers. Despite the emergence of new compressionalgorithms capable of providing extremely high compression ratios, there is still a challenge to provide optimised storage services capable of storing 1000s of hours of multimedia data and providing simultaneous access to hundreds and potentially thousands of clients. A number of prototype multimedia storage devices have been designed and built to address the real-time demands of digitalvideo and audio. These systems use RAID (Redundant Arrays of Inexpensive Disks) technology to increase the bandwidth and storage capacities of single disks. Unfortunately, RAID is non-scalable (i.e. disks cannot be incrementally added to provide higher bandwidth and storage facilities) and can prove to be a bottleneck in the system. This servo-bottleneck only becomes significant when the aggregate disk bandwidth of RAID increases beyond either the network interface or the server CPU / bus speed. Once this occurs the available storage and bandwidth can only be increased by introducing more RAID servers and introducing a load balancing policy where client requests can be shared across multiple server instances. This paper describes a Scalable Multimedia Storage Architecture (SMSA) that supports wide area storage, storage server scalability allowing the addition of extra storage nodes, and maximised available data streams through the use of the load balancing techniques of network striping / file replication. It also allows for the storage of multi-resolution data produced by scalable compression techniques to match the Quality of Service requirements of heterogeneous clients.
This paper describes a possible strategy to migrate from block-based coding towards object-based coding. We use the results of an intelligent motion estimation algorithm to define low resolution (block based) segmenta...
详细信息
ISBN:
(纸本)0819423564
This paper describes a possible strategy to migrate from block-based coding towards object-based coding. We use the results of an intelligent motion estimation algorithm to define low resolution (block based) segmentation. The information from previous segmentation results together with the motion vector field, the image data and spatial relations are used to define the cost to belong to a given object. A separate selection procedure is used to generate the high resolution (pixel based) objects. Afterwards some shape simplification is performed to generate the final objects.
This paper describes the three main international video-compression standards and reviews several industry-standard video-compressionalgorithms to compare their strengths, weaknesses, and intended applications. Some ...
详细信息
This paper describes the three main international video-compression standards and reviews several industry-standard video-compressionalgorithms to compare their strengths, weaknesses, and intended applications. Some related standard under development are also studied.
An efficient algorithm for dynamically multiplexing MPEG2 encoded video sources is presented. Sources are grouped into classes regarding different combined levels of spatial detail and amount of movement. Simulations ...
详细信息
ISBN:
(纸本)0819423564
An efficient algorithm for dynamically multiplexing MPEG2 encoded video sources is presented. Sources are grouped into classes regarding different combined levels of spatial detail and amount of movement. Simulations were performed using different associations of sources belonging to distinct classes, different bit rates and GOP structures. The implications associated to a real implementation are analyzed and a modular architecture is proposed. Simulation results are presented and discussed, showing that sequences with higher spatial detail and motion are those which exhibit the higher quality improvements. These results are almost not affected by the non alignment, at GOP level, between video sequences.
Until recently, video communication was almost exclusively reserved for the all-digital ISDN world because it requires more than four-and-a-half times the bandwidth of analog telephone networks. But new standards and ...
详细信息
Until recently, video communication was almost exclusively reserved for the all-digital ISDN world because it requires more than four-and-a-half times the bandwidth of analog telephone networks. But new standards and compression techniques have now made video communication feasible in the analog world too. Complex audio and videocompressionalgorithms in particular have opened the door to video communication in analog networks. Together, JADE AN for analog networks and JADE MM with combined analog and digital capability will make worldwide video communication to standards H.320 and H.324 possible for the first time ever.
This paper describes the technology used in a new generation of digital cameras. The cameras all use full-frame image sensors optimized for producing still images in electronic cameras. The color CCDs incorporate the ...
详细信息
ISBN:
(纸本)081942028X
This paper describes the technology used in a new generation of digital cameras. The cameras all use full-frame image sensors optimized for producing still images in electronic cameras. The color CCDs incorporate the Bayer color filter array pattern, lateral overflow drain antiblooming protection, accumulation mode timing, and progressive scan readout. The three cameras described have CCDs with 6.3 million, 1.6 million, and 400 thousand pixels, with 9 micron square pixels and a 3:2 image aspect ratio. They use a firmware-based digital camera architecture to maximize flexibility and image quality. Extensive digital image processing is performed in the host computer, as the images are downloaded from the camera. This enables the cameras to use sophisticated image processing algorithms that can be easily upgraded in the field and customized for special customer applications.
Progress of Optical Information Processing is driven by two sources: optoelectronic technology benefits from the development of compound semi-conductor circuit technology, and the continued growth of electronic chip s...
详细信息
Progress of Optical Information Processing is driven by two sources: optoelectronic technology benefits from the development of compound semi-conductor circuit technology, and the continued growth of electronic chip speed and density increases the need for new interconnect technologies. We advocate the use of these advances for solving complex low level image processing problems using dedicated optoelectronic processor arrays. These massively parallel, special purpose machines would not be conceivable without a suitable combination of optics with high density hybrid (digital and analog, electronic and optical) smart pixels. One optical operation of particular interest here is the generation of random number arrays for the implementation of stochastic algorithms. We describe the general structure of optoelectronic `smart retinae' that combine laser speckle, optical convolution and VLSI circuits to implement parallel simulated annealing. We report on our present results that have demonstrated one simple case of such an operation performing parallel simulated annealing at video rate.
暂无评论