Title :
Multiprocessor computing for images
Author :
Cantoni, Virginio ; Levialdi, Stefano
Author_Institution :
Dept. Inf. e Syst., Pavia Univ., Italy
fDate :
8/1/1988 12:00:00 AM
Abstract :
A review of image-processing systems developed until now is given, highlighting the weak points of such systems and the trends that have dictated their evolution through the years producing different generations of machines. Each generation may be characterized by the hardware architecture, the programmability features and the relative application areas. The need for multiprocessing hierarchical systems is discussed, focusing on pyramidal architectures. Their computational paradigms, their virtual and physical implementation, their programming and languages, are discussed
Keywords :
computer architecture; computer vision; multiprocessing systems; application areas; computational paradigms; computer vision; evolution; hardware architecture; image-processing systems; multiprocessing hierarchical systems; physical implementation; programmability features; programming; pyramidal architectures; review; trends; weak points; Application software; Character generation; Computer architecture; Computer languages; Hardware; Hierarchical systems; Image generation; Image processing; Image segmentation; Image storage; Physics computing;
Journal_Title :
Proceedings of the IEEE