Fuzzy C-Means With Local Membership Based Weighted Pixel Distance and KL Divergence for Image Segmentation
The Journal of Pattern Recognition Research (JPRR) provides an international forum for the electronic publication of high-quality research and industrial experience articles in all areas of pattern recognition, machine learning, and artificial intelligence. JPRR is committed to rigorous yet rapid reviewing. Final versions are published electronically
(ISSN 1558-884X) immediately upon acceptance.
Fuzzy C-Means With Local Membership Based Weighted Pixel Distance and KL Divergence for Image Segmentation
Reda R Gharieb, G. Gendy
JPRR Vol 10, No 1 (2015); doi:10.13176/11.605 
Download
Reda R Gharieb, G. Gendy
Abstract
This paper presents a new technique for incorporating local membership information into the standard fuzzy C-means (FCM) clustering algorithm. In this technique, the objective consists of minimizing the classical FCM function with unity fuzzification exponent plus the Kullback–Leibler (KL) information distance acting as a fuzzification and regularization term. The KL distance is proposed to measure the proximity between cluster membership function of a pixel and an average of the cluster membership functions of immediate neighborhood pixels. Therefore, minimizing this KL distance biases the cluster membership of the pixel toward this smoothed membership function of the local neighborhoods. This can provide immunity against noise and results in clustered images with piece-wise homogeneous regions. Results of clustering and segmentation of synthetic and real-world medical images are presented to compare the performance of the proposed local membership KL information based FCM (LMKLFCM) and the standard FCM, a local data information based FCM (LDFCM) and a type of local membership information based FCM (LMFCM) algorithms.
JPRR Vol 10, No 1 (2015); doi:10.13176/11.605 | Full Text  | Share this paper: