International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 119 - Number 8 |
Year of Publication: 2015 |
Authors: Utkaleswar Padhan, Sagar Kumar Tripathy, Sudhakar Sahoo, Birendra Kumar Nayak, Om Prakash Jena |
![]() |
Utkaleswar Padhan, Sagar Kumar Tripathy, Sudhakar Sahoo, Birendra Kumar Nayak, Om Prakash Jena . Sweeper's Algorithm and its Application on Image Clustering. International Journal of Computer Applications. 119, 8 ( June 2015), 30-37. DOI=10.5120/21090-3785
Using Two Dimensional Hybrid Cellular Automata (2-D HCA) rules image shifting, image copying, zooming in and out, thickening and thinning of an image etc. were possible and reported in [8]. A new searching algorithm called Sweeper’s algorithm was proposed on binary images using 2-D HCA and was the basis for solving various problems like migration of initially distributed organisms in a space towards a single point destination, density classification problem of CA etc. [9]. Due to the wide scope of Sweeper’s algorithm it was mentioned in [8] that the problem areas that can be solved by this algorithm are text and image compression, informed search in Artificial Intelligence, clustering problem, Cryptography, and pattern classification etc. Here in this paper we have taken up the challenge and tried to solve the clustering problem using Sweeper’s algorithm. Basically our study is based on various color images consisting of the combinations of Red, Green, and Blue (RGB) colors as well as selecting different destination points in the search space. First we have taken red color and one of the two color from blue or green as cluster points and took it as input and by applying Sweeper’s algorithm we found two cluster regions one with red color and other with blue or green color. Subsequently we took all the three (RGB) color as input and by applying the algorithm we found three clusters. We took one color between red, blue, green and some other color like magenta, cyan, yellow etc. as input and found different clusters. We have also studied the intersection region of different colors using this algorithm and found interesting color patterns.