VLSI (Very Large Scale Integration) Architectures for Pattern Analysis and Image Database Management
In this presentation, VLSI computing structures are introduced for the analysis and management of imagery data and both pattern-analysis and image-database-management capabilities into a unified system design.
Abstract
Abstract : In this presentation, VLSI computing structures are introduced for the analysis and management of imagery data. Information scientists have long recognized the fact that one picture is worth a thousand words. Machine intelligence would be greatly enhanced if new generation of computers could be designed to process multidimensional imagery data above the string processisng of alphanumerical information by present computers. Over the last decade, extensive research and development has challenged to achieve the capabilities of pattern analysis and image understanding by computers. Practical applications of such computers include the processing of biomedical images for diagnosis, the recognition of characters, fingerprints, and moving objects, remote sensing, industrial inspection, robotic vision, military intelligence, and data compression for communications. Image analysis refers to the use of digital computers for Pattern Recognition and Image Processing (PRIP). On-line imagery data needs to be restored on disks and fastly retrieved for PRIP applications. A VLSI-based image analysis machine should integrate both pattern-analysis and image-database-management capabilities into a unified system design. (Author)