Parallel solution of Hough transform and convolution problems—a novel multimodal approach
Lan Jin, Lan Yang · 1992
In this paper, we proposed new message-passing parallel algorithms for solving Hough transform and convolution problems on a multimodal architecture.The algorithms were designed and implemented on the basis of three different computation models, namely, geometric parallelism, algorithmic parallelism, and processor farming.Proper choice of computation models helps fully exploit parallelism inherent in various application problems, The algoridurrs take full advamagc of the flexibility of the rcconfigurable multimodal rrrchitechrrc of our transputer-based mukicomputer system.In result, our solution effectively balances such opposing factors, as time vs. resource, processing vs. communications, etc.The novel approach propased in this paper to the design and implementation of parallel algorithms based on different computation models has enabled us to realize simple, flexible and cfficicnt solution of Lhe two basic problems in image and signal processing... .. .. ..