In this paper, a parallel and distributed algorithm on Spiral Architecture for edge detection using a star network is proposed. The proposed algorithm is based on Master-Slave model. However, it guarantees a better load balancing among the processors (or nodes) and greatly reduces the traffic from and to the master node. It also requires less processes be created on the master node and hence saves CPU time for less process switches. Our work sees a potential for further speeding up distributed image processing such as object recognition and pattern recognition based on computer clusters.
|Title of host publication||3rd IASTED International Conference on Visualization, Imaging, and Image Processing, VIIP 2003|
|Place of Publication||Canada|
|Publication status||Published - 2003|
|Event||International Conference on Visualization, Imaging and Image Processing - Benalmádena, Spain, Spain|
Duration: 08 Sep 2003 → 10 Sep 2003
|Conference||International Conference on Visualization, Imaging and Image Processing|
|Period||08/09/03 → 10/09/03|
He, X., Wu, Q., Liu, D., & Zheng, L. (2003). A Distributed and Parallel Edge Detection Scheme within Spiral Architecture. In M. H. Hamza (Ed.), 3rd IASTED International Conference on Visualization, Imaging, and Image Processing, VIIP 2003 (pp. 396-303). ACTA Press.