A design methodology for the definition of parallel architectures for image processing is presented. The architectural design is based on functional blocks, operating on limited regions of neighboring pixels (windows) in the image plane. This window-based computational granularity allows the coverage of a wide range of image processing algorithms, with particular reference to image enhancement and analysis. The window-based formulation and decomposition of the algorithms is derived, making possible the identification of the architectural characteristics of the window processor and of the required interconnection structures. Evaluations of the architectural complexity guaranteed by this approach are given in terms of temporal and spatial requirements.
Window-based functional blocks for image processing
ANTOLA, ANNA MARIA;BREVEGLIERI, LUCA ODDONE
1991-01-01
Abstract
A design methodology for the definition of parallel architectures for image processing is presented. The architectural design is based on functional blocks, operating on limited regions of neighboring pixels (windows) in the image plane. This window-based computational granularity allows the coverage of a wide range of image processing algorithms, with particular reference to image enhancement and analysis. The window-based formulation and decomposition of the algorithms is derived, making possible the identification of the architectural characteristics of the window processor and of the required interconnection structures. Evaluations of the architectural complexity guaranteed by this approach are given in terms of temporal and spatial requirements.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
