Abstract: Image segmentation plays a crucial role in many image applications. Fuzzy C-Means (FCM) is a powerful method to segment image and has been applied to many image segmentations successfully. However, on the one hand, traditional FCM algorithm is sensitive to the noise due to the fact that it only accounts for the pixel’s value information, not takes the neighboring pixels’ spatial information into consideration; on the other hand, its performance becomes poor when the input data are nonlinearly. To overcome the above mentioned problems, this study proposed a neighboring weighted FCM associated with kernel method algorithm (KNW-FCM) for image segmentation. The values of the coefficients within neighboring window are adaptively determined by the characteristics of the image itself. Furthermore, the original Euclidean distance in the traditional FCM is replaced by the kernel-induced distance. Additionally, as a real application, we apply the proposed algorithm to segment the microscopic image of harmful algae. Experimental results show that the proposed algorithm achieves better performance compared with other algorithms.