Shen K. and Delp E. J., "A Spatial-Temporal Parallel Approach For Real-Time MPEG Video Compression, Proceedings of the 25th International Conference on Parallel Processing, pp. II100-II107  Home/Search   Document Details and Download   Summary   Related Articles   Check

....1 k DCI for all I frames contained in V. A number of different norms [5 7] are available for calculating the differences. Many of these are very compute intensive. The norm proposed here requires relatively little compute effort and has been tested with success in the case of compressed videos [9 11]. If a scene cut exists between two I frames the calculated distance between the two must be large compared to the average distance between I frames. In practice a scene cut spans a number of I frames and it is thus reasonable to assume that within a sequence of difference images a large value ....

K. Shen and E. Delp, A spatial-temporal parallel approach for real-time MPEG Video compression, Image Compression and Graphics, pp. 100-107, 1996.

.... like digital library or production systems which require encoding a video sequence and storing it on a CD ROM or DVD (Digital Versatile Disk) There are two approaches to performing video compression: hardware based [1] 23] and software based [2] 3] 4] 7] 10] 11] 19] 20] [22], 24] Both approaches have their own advantages and disadvantages. A hardware approach uses a special purpose architecture, and its advantages include the ease of use and high compression speed. However, dedicated hardware is less flexible and can become obsolete. Furthermore, hardware is often ....

....encoding a large video sequence into a disk or CD ROM. In this approach, one can partition the video data in the temporal domain and let each processors independently encode a sequence of frames, such as a GOP. The pioneering work done for non real time environment has been reported in [21] and [22]. In [21] GOP level temporal parallelism has been used for MPEG1 using the Intel Touchstone Delta and the Intel Paragon. I O contention has been found to be the major bottleneck, and an I O management policy has been used to restrict the number of PEs (processing elements) that involve in I O at ....

[Article contains additional citation context not shown here]

Ke Shen and Edward Delp, A Spatial-Temporal Parallel Approach For RealTime MPEG Video Compression, in Proc. of 1996 Intl. Conf. on Parallel Processing , 2 (1996), 100-107.

....video. When image size is large or high image quality is necessary with low data volume, video compression in real time is difficult to attain and very expensive. For this reason, parallel processing becomes the natural choice to satisfy the computational demands of video compression in real time [1]. This work was partially supported by CICYT project TIC97 0897 C04 02 A video sequence can be seen as a threedimensional signal (3 D) two dimensions in the spatial domain and one in the temporal domain. Thus, two different approaches can be used to parallelize the codification algorithm, ....

Shen, K. Delp, E.J. "A spatial-Temporal Parallel approach for Real-Time MPEG Video Compression". Universidad de Purdue. School of Electrical Engineering Computer Vision and Image Processing Laboratory and Purdue multimedia Testbed. West Lafayette, Indiana 47907-1285.

....switch and also via Ethernet. We exploit parallelism on a Group of Picture (GOP) basis such that each GOP of the video is encoded by a particular processor. We propose a scheme for efficient I O and data distribution. 1 Introduction Video compression can be hardware based [1] or software based [2, 3]. However, a software solution is more flexible, and thus allows algorithmic improvements. Nonetheless, very high computation requirements of video applications can often overwhelm a single processor sequential computer [2] Therefore, it is natural to exploit the enormous computing power offered ....

K. Shen, L. A. Rowe, E. J. Delp, "A Spatial-Temporal Parallel Approach for Realtime MPEG Video Compression", Proc. of 1996 International Conf. on Parallel Processing, vol. 2, 1996, pp. 100-107.

....real time TV broadcasting. These devices are very expensive, e.g. a real time MPEG2 encoder 9for ITU R 601 video costs in excess of 25,000. In many cases only a single video stream can be processed at a time. Software approaches include the use of parallel computers and networks of workstations [27, 28, 17, 29, 13, 14, 30, 31]. Usually a software implementation o#ers more flexibility, e.g. parameters can be adjusted easily and multiple video streams can be handled, which is essential for applications such as a video server or a video database. Portability may also be an advantage of a software implementation. A ....

....compression speed (execution time) the time needed for the I O operations was not taken into account. When I O operations are included, the algorithm will not obtain real time performance. In temporal parallelism, di#erent PEs are assigned to process di#erent video frames of a video sequence [17, 13, 31]. There is no upper limit on how many PEs can be used since a typical 2 hour motion picture contains more than 200,000 frames. A problem with this approach is that even though the overall performance (throughput) can be faster than real time the compression of a single frame may be slower than ....

[Article contains additional citation context not shown here]

K. Shen and E. J. Delp, "A spatial-temporal parallel approach for real-time mpeg video compression," Proceedings of the 25th International Conference on Parallel Processing, August 13--15, 1996, Bloomingdale, Illinois, pp. II--100--II-- 107.

....is that even though the overall performance (throughput) can be faster than real time the compression of a single frame may be slower than real time, which results in a constant delay. A variation of the above approaches, which combines the temporal and spatial parallelism, has been proposed [23]. In this paper, we describe the spatial temporal parallel approach in detail. New experimental results and an analysis of the execution time of each task module of the algorithm are presented. A comparison of this method with other parallel approaches are also given. The rest of the paper is ....

....Nodes and the one located at Purdue University which has 140 Numeric Nodes. The total computation capacities exceed 38 Gflops and 10 Gflops, respectively. Both systems have multiple PFSs. 4. Problem Description Several approaches have been proposed to map the MPEG algorithm on to the Intel Paragon [1, 22, 23]. The main objective of the work has been to develop parallel implementations that provide real time throughput. In our experiments, we found that a single node on the Paragon can compress a CCIR 601 video sequence at a speed of 0.09 frames per second with 95.4 of the time used for computation. ....

[Article contains additional citation context not shown here]

K. Shen and E. J. Delp. A spatial-temporal parallel approach for real-time mpeg video compression. In Proceedings of the 25th International Conference on Parallel Processing, pages II--100--II--107, Bloomingdale, Illinois, August 13--15, 1996.

No context found.

Shen K. and Delp E. J., "A Spatial-Temporal Parallel Approach For Real-Time MPEG Video Compression, Proceedings of the 25th International Conference on Parallel Processing, pp. II100-II107

No context found.

K. Shen and E.J. Delp. A spatial-temporal parallel approach for real-time mpeg video compression. Proceedings of the 1996.

No context found.

K. Shen and E. Delp, A spatial-temporal parallel approach for real-time MPEG Video compression, Image Compression and Graphics, pp. 100-107, 1996.

No context found.

K. Shen, E. Delp, A spatial-temporal parallel approach for real-time MPEG Video compression, Image Compression and Graphics, 100-107, 1996.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC