Iris Recognition Modality with Fuzzy k-Means Clustering Algorithm for Improving Security

Abstract

This paper discusses improving security systems using an iris recognition modality with a new technique called fuzzy k-means. Existing fuzzy c-means will work based on the iteration. If the number of iterations increases then it will take more processing time, and also the KNN classifier works at a high degree of local sensitivity which will affect choosing the training set. Due to the lack of robustness in the KNN classifier and time complexity in the FCM algorithm a new method has been proposed called the IRIS model with fuzzy k-means clustering algorithm and feature extraction using principal component analysis (PCA) and hidden Markov model clustering (HMM). Here the fuzzy k-means are used for clustering that provide the data clustering with less time and higher segmentation efficiency compared to fuzzy clustering. It will segment features such as sclera, pupils, and clustered skin in the iris image and give the features of the pupils. PCA will fetch the data with lower dimensionality and this lower dimensionality data can be used for extracting features from the clustered image. A hidden Markov model can be used to classify the features from the clustering. HMM is good at dealing with sequential inputs. It provides the improved classification rate with higher accuracy.