Fuzzy C-Means with Membership Constraints Using Kernel-Induced Distance Measure and its Applications on Infrared Image Segmentation

Image segmentation plays a crucial role in many fields. In this paper, we present a novel algorithm for fuzzy segmentation of infrared imaging data. The algorithm is realized by modifying the objective function in the fuzzy C-means with improved fuzzy partition(FCM-IFP) using a kernel-induced distance metric, namely, the original Euclidean distance in the FCM-IFP is replaced by a kernel-induced distance, and thus the corresponding algorithm is derived and called as the kernelized FCM-IFP (KFCM-IFP). This processing method not only can suppress the noise and the outliers, but also can prevent the over segmentation of infrared image even if the contrast between targets and background is insufficient. The experimental results show that the infrared image can be segmented well by the proposed method compared with the conventional clustering method, and the noise, outliers and insufficient contrast are prevented to influence the segmentation of targets region.