Title :
Concurrent edge detection with Spiral Architecture on Linux
Author :
Hu, Qimei ; He, Xiangjian ; Wu, Qiang
Author_Institution :
Dept. of Comput. Syst., Univ. of Technol., Sydney, NSW, Australia
Abstract :
Edge detection is an essential tool in image analysis. It is a process of identifying significant discontinuities in light intensities in order to form an outline of the object of interest. Spiral Architecture provides powerful computational power as a method of image data representation. Gaussian multi-scale theory was employed as the mathematical model for edge detection. The parallel processing algorithm was implemented in the form of a master-slave model. Spiral Architecture enables an image to be uniformly partitioned and distributed to slave processors after spiral multiplication. Concurrent processing is facilitated by inter-process communication and socket interface programming tools of Linux operating system.
Keywords :
Unix; edge detection; image representation; message passing; parallel algorithms; Gaussian multi-scale theory; Linux operating system; Spiral Architecture; concurrent edge detection; concurrent processing; image analysis; image data representation; image partitioning; inter-process communication; light intensities; master-slave model; parallel processing algorithm; slave processors; socket interface programming tools; spiral multiplication; Computer architecture; Image edge detection; Linux; Master-slave; Mathematical model; Parallel processing; Partitioning algorithms; Power system modeling; Sockets; Spirals;
Conference_Titel :
Information Technology: Coding and Computing [Computers and Communications], 2003. Proceedings. ITCC 2003. International Conference on
Print_ISBN :
0-7695-1916-4
DOI :
10.1109/ITCC.2003.1197584
