摘要

Image segmentation aims to separate the structure of interest objects from the background and the other objects. Many approaches have been developed to segmentation of brain MR images, which among them the fuzzy c-mean (FCM) algorithm is widely used in MR images segmentation. In this paper we introduce a noise probability for each pixel within the image; then according to their noise probability, the artificial bee colony (ABC) algorithm classified all pixels into two groups: Normal and Noisy. In proposed algorithm, ABC optimization is used before performing the FCM clustering algorithm. Our algorithm reduced the response time with a higher quality than the previous approaches. Finally, we segment some real MR images with the proposed algorithm and compare it with the other approaches.

  • 出版日期2011