A variant of fuzzy c-means (FCM) clustering algorithm for image segmentation is provided. Unlike the L2-norm distance in FCM, Lp with p ∈ (0,1] norm is used to measure the distance of the pixel intensity to its cluster centre in the energy functional. Moreover, local spatial information and colour information are incorporated into the model to enhance the robustness to noise and outliers. The proposed algorithm is called fuzzy local information Lp (FLILp) clustering. To overcome the difficulty of finding cluster centres, Lp-norm distance is approximated by weighted L2 distance. The advantages of FLILp are: (i) it is strongly robust to noise and outliers, (ii) it is applied to the original image and (iii) it preserves image edges. Numerical examples and comparisons of image segmentation on both synthetic and real images illustrate the outstanding performance and robustness of the proposed method.
|Number of pages||10|
|Journal||IET Image Processing|
|State||Published - Apr 1 2017|
Bibliographical noteFunding Information:
The research of F. Li is supported by the National Science Foundation of China (No. 11671002) and the Science and Technology Commission of Shanghai Municipality (STCSM) (No. 13dz2260400). The research of Jing Qin is supported by the faculty start-up fund of Montana State University
© The Institution of Engineering and Technology.
ASJC Scopus subject areas
- Signal Processing
- Computer Vision and Pattern Recognition
- Electrical and Electronic Engineering