Iris is the sphincter having flowery pattern around pupil in eye. The high randomness of the pattern makes iris unique for each individual, and hence a candidate for machine recognition of identity of an individual. The first part of the thesis investigates the bottlenecks of the existing localization approaches and proposes a morphological method of devising an adaptive binarization threshold for pupil detection, and also contributes in modifying conventional integrodifferential operator based iris detection using canny detected edge map. The review of related works on matching leads to the observation that local features like Scale Invariant Feature Transform(SIFT) matches the keypoints on the basis of 128-D local descriptors, hence it sometimes falsely pairs two keypoints which are from different portions of two iris images. Subsequently the need for pruning of faulty SIFT pairs is felt. The second part of the thesis proposes two methods of filtering (Angular Filtering and Scale Filtering) the SIFT impairments (faulty pairs) based on the knowledge of spatial information of the keypoints. The pruning algorithms experimentally show higher accuracy and increased separability.
