Search

In this work, we proposed a method for fingerprint image segmentation based on possibilistic fuzzy c-means (PFCM) algorithm with an adaptive level set (ALS) method. In fingerprint recognition system, fingerprint segmentation is an important step. PFCM algorithm is a mixer of possibilistic c-means clustering (PCM) and fuzzy c-means clustering (FCM) algorithm. PFCM overcomes the noise sensitivity defect in FCM and coincident cluster problem in PCM. PFCM was used to generate an initial contour curve for level set method. PFCM algorithm is used to compute the fuzzy membership values of each pixel. Based on the above fuzzy membership values edge indicator function is redefined. By using the edge indicator function fingerprint segmentation was performed to extract the required regions for advance processing. Experimental results of proposed method showed significant improvement in the evolution of the level set function.

Image segmentation refers to the process of partitioning an image into its constituent parts or objects. The identical images can be segmented manually, but the accurateness of image segmentation using the segmentation algorithms is more compared to manual segmentation. Clustering is a primary data description method in applications such as data mining, pattern recognition and bioinformatics. The data clustering is an important problem in image processing. Different algorithms are proposed to solve this clustering problem. In this paper, we proposed a multiple kernel fuzzy c-means (MKFCM) on level set method. MKFCM was used to generate an initial contour curve which overcomes leaking at the boundary during the curve propagation. This multiple kernel fuzzy c-means provides us to combine different information of image pixels in segmentation algorithms. The different types of information of image pixels are combined in the kernel space by combining different kernel functions defined on specific information domains. Based on MKFCM the edge indicator function was redefined. Using the edge indicator function image segmentation was performed. The efficiency and accuracy of the proposed algorithm is shown by experimenting on the MRI brain image and satellite images.