基于FCM算法与互信息量的图像自动分割

传统的阈值分割算法只考虑到图像的灰度信息,而忽略了灰度的空间分布以及分割后图像与原图像之间的关系。本文从分割图像与原图像的内在联系出发,提出了一种新的基于FCM算法与互信息量技术相结合的分割算法,即FCM-MI算法。首先利用FCM算法确定全局阈值作为初值,以互信息量为目标函数,在小范围内计算分割图像与原图像的互信息量,互互信息量达到最大时的阈值即为最优值。对大量医学图像和车牌图像进行的实验结果表明,本算法所得到的目标图像的边界特征保持完好,虚假目标信息大大降低,图像边界细腻、连续且定位性能好。

Unsupervised Segmentation of Images Based on FCM and Mutual Information

Most image segmentation algorithms rely on statistical methods, without taking the relationships between the pixels into account. In the paper,the FCM algorithm is combined with the mutual information (MI) technique. The initial threshold can be chosen by using the FCM algorithm, and in the iteration process, an optimal threshold will be determined by maximizing the MI between the original and segmented images. We evaluate the effectiveness of the proposed approach by applying it to medical images and license plate images. The experimental results indicate that the proposed method has not only visually better or comparable segmentation effect but also, more favorably,the removal ability for noises.