摘要 |
PROBLEM TO BE SOLVED: To improve the processing performance of edit and search operations for long and huge continuous media data by parallel data processing. SOLUTION: In this method, each of two or more continuous media data, which are inputted in parallel, is sequentially divided at a specified interval in time base to generate data blocks. The data blocks are grouped in the unit of continuous media data to generate block sets, and then they are stored in a storage device. A read processing to read each block set from each storage device, a comparison processing to perform specified comparison for each of the read block sets, and data transfer processing to transfer data of each comparison result are sequentially performed. These processings are performed for the plurality of block sets asynchronously, in simultaneous parallel, according to the number of block sets. COPYRIGHT: (C)2003,JPO
|