Neighbourhood weighted fuzzy c-means clustering algorithm for image segmentation

• Published in: IET image processing Vol. 8; no. 3; pp. 150 - 161
• Main Authors: Zaixin, Zhao, Lizhi, Cheng, Guangquan, Cheng
• Format: Journal Article
• Language: English
• Published: Stevenage The Institution of Engineering and Technology 01.03.2014; Institution of Engineering and Technology; The Institution of Engineering & Technology
• Subjects: Applied sciences; Biological and medical sciences; Computerized, statistical medical data processing and models in biomedicine; Euclidean distance replacement; Exact sciences and technology; fuzzy set theory; Image processing; image segmentation; Information, signal and communications theory; local contextual information; Medical management aid. Diagnosis aid; Medical sciences; modified FCM algorithm; neighbourhood weighted distance; neighbourhood weighted fuzzy c‐means clustering algorithm; objective function; Signal and communications theory; Signal processing; Signal representation. Spectral analysis; Signal, noise; structure information; Telecommunications and information theory; neighbourhood weighted fuzzy c-means clustering algorithm; image segmentation; fuzzy set theory; modified FCM algorithm; structure information; neighbourhood weighted distance; local contextual information; Euclidean distance replacement; objective function; Image segmentation; Automatic classification; Image processing; Similarity; Medical imagery; Artefact; Objective function; Information structure; Algorithm; Signal classification; Nuclear magnetic resonance imaging
• ISSN: 1751-9659; 1751-9667; 1751-9667
• DOI: 10.1049/iet-ipr.2011.0128

Fuzzy c-means (FCM) clustering algorithm has been widely used in image segmentation. In this study, a modified FCM algorithm is presented by utilising local contextual information and structure information. The authors first establish a novel similarity measure model based on image patches and local statistics, and then define the neighbourhood-weighted distance to replace the Euclidean distance in the objective function of FCM. Validation studies are performed on synthetic and real-world images with different noises, as well as magnetic resonance brain images. Experimental results show that the proposed method is very robust to noise and other image artefacts.