An Effective and Efficient FCM for Segmenting Color Image with High Intensity Noise

DOI

10.2991/icismme-15.2015.343How to use a DOI?

Keywords

fuzzy c-means clustering; fast clustering; alpha-trimmed; image segmentation.

Abstract

In this paper, an effective and efficient fuzzy c-means algorithm is proposed for color image segmentation. To make the segmentation robust, the pixel similarity, the sliding window and alpha-trimmed mean are combined together to suppress the influence of heavy noise during the segmentation process. On the other hand, to make the segmentation effective, the quantitative technique is utilized in RGB space to accelerate the speed of segmentation. Our algorithm has two main advantages: (1) By considering the local spatial and color information together, the high intensity noise can be removed effectively to great extent; (2) Due to the using of quantitative techniques in RGB spaces, the time complexity of our algorithm is proportional to the number of colors (only 216) and independent of the number image pixels, and therefore it is very suitable for large-scale images. The experiments show that our algorithm is effective and efficient.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Weiling Cai
AU  - Wang Li
PY  - 2015/07
DA  - 2015/07
TI  - An Effective and Efficient FCM for Segmenting Color Image with High Intensity Noise
BT  - Proceedings of the First International Conference on Information Sciences, Machinery, Materials and Energy
PB  - Atlantis Press
SP  - 1653
EP  - 1659
SN  - 1951-6851
UR  - https://doi.org/10.2991/icismme-15.2015.343
DO  - 10.2991/icismme-15.2015.343
ID  - Cai2015/07
ER  -