Title :
Image segmentation by fuzzy rule and Kohonen-constraint satisfaction fuzzy C-mean
Author :
Khunkay, Suraphon ; Paithoonwattanakij, Kitti
Author_Institution :
Dept. of Electron., King Mongkut´´s Inst. of Technol., Bangkok, Thailand
Abstract :
We present a new algorithm that can segment fuzzy data. This method is based on fuzzy logic and neural network, and the concept of the constraint satisfaction problem (CSP). Firstly, pre-processing by creating a new pixel value based on the fuzzy rule has the ability to describe the effect of the neighborhood pixel on the degree of membership value or linguistic variables which are utilized to activate a rule base. Secondly, the feature extraction employs Kohonen (1982) feature mapping (SOM), which can learn a feature without prior knowledge. Finally, a fuzzy C-mean (FCM) adapts to the CSP structure by constituting a interconnection weight for all membership values of the fuzzy c-mean with a global consistency situation. The result of this method have been tested on a variety of images
Keywords :
feature extraction; fuzzy logic; image segmentation; knowledge based systems; self-organising feature maps; unsupervised learning; Kohonen self organizing feature map; Kohonen-constraint satisfaction; constraint satisfaction problem; feature extraction; fuzzy C-mean; fuzzy data segmentation; fuzzy logic; image segmentation; interconnection weight; linguistic variables; membership value; neural network; pixel value adjustment; preprocessing; rule base; unsupervised learning neural network; Data preprocessing; Feature extraction; Fuzzy logic; Fuzzy sets; Image segmentation; Neural networks; Organizing; Pixel; Testing; Unsupervised learning;
Conference_Titel :
Information, Communications and Signal Processing, 1997. ICICS., Proceedings of 1997 International Conference on
Print_ISBN :
0-7803-3676-3
DOI :
10.1109/ICICS.1997.652070
