Support Vector Machine Based Iris Recognition System for Personal Identification

Abstract: Iris recognition is emerging as important method of biometrics-based identification systems. Image preprocessing is performing at first, follow by extracting the iris portion of the eye image. The extracted iris part is then normalize and then Iris Code is construct using 1D gabor filter.The selection of the finest feature subset and the categorization has grow to be an important issue in the field of iris recognition. In this paper we propose several methods for iris feature subset selection and vector creation. The acceptable feature sequence is extracted from the iris image by using the contour let transform technique. Contour let transform capture the fundamental geometrical structures of iris image. It decompose the iris image into a set of directional sub-bands with texture particulars captured in different orientations at a variety of scales so for dropping the feature vector dimensions we use the method for extract only significant bit and information from normalized iris images. In this method we disregard fragile bits and finally we use SVM (Support Vector Machine) classifier for resembling the amount of people identification in our proposed system. Experimental result show that most projected method reduces processing time and enlarge the classification accuracy and also the iris feature vector length is much smaller versus the other method. Experimental image results illustrate that, exclusive codes can be generated for every eye image. Iris recognition analysis the features that exist in the colored tissue neighboring the pupil, comparison, rings , furrows and freckles.

Keywords: Biometric-Iris Recognition, Contourlet-Support VectorMachine (SVM), Grey Level co-occurrence matrix(GLCM), IRIS.