ix SEGMENTATION IMPLEMENTATION OF IMAGES COMPRESSED BY FUZZY TRANSFORMS USING THE FAST GENERALIZED FUZZY C-MEANS ALGORITHM

In this Final Project, Fast Generalized Fuzzy C-Means (FGFCM) algorithm has been used for segmenting an images compressed by fuzzy transforms algorithms. The images compressed input is needed for solving the problem about size of dataset, before the dataset implemented in segmentation process that used spacial information and gray value based on correlation between pixels. There are several steps in this Final Project. The first step is image compressing by using direct fuzzy transforms algorithm. The second step is restoring this image to its original size (decompress) using invers fuzzy transforms. And the final step is segmenting the compressed image using FGFCM algorithm to get parts of processed images. Experimental result has been proved that FGFCM algorithm with images compressed by fuzzy transforms can produce the segmentation time better than segmentation comparity to original images without compressing. And it can produce dimension of the image dataset smaller too. Besides that the size of compression also affect the results of segmentation where the size of the greater compression will produce the better of segmentation too, it can be seen based on PSNR values obtained from each process.