Abstract: Traditional image threshold segmentation algorithm in MR image segmentation is easy to be disturbed by the problem of gray un-uniform and being susceptible to noise interference. Considering these problems, an MR image segmentation algorithm based on artificial bee colony optimization is proposed, which is combining the artificial bee colony algorithm with the two-dimensional OSTU threshold segmentation algorithm. The algorithm uses the trace of the dispersion matrix of medical image as the objective function of artificial bee colony optimization, the optimal segmentation threshold of two-dimensional OSTU is obtained. According to the optimal threshold, the image is segmented by two-dimensional OSTU. The experimental results show that, for medical MR images, the algorithm proposed in this paper has the characteristics of high accuracy and strong robustness, and can get accurate segmented images.

Key words: artificial bee colony, 2D OSTU, MR image