A wavelet-based feature extraction technique for partial iris recognition

Feature extraction is an important task in the overall processing of iris biometric in an iris based biometric authentication system. The existing approaches are using complete iris image to extract iris features. But, it is difficult to get complete iris image for feature extraction because a person's eyes are naturally covered by the eyelids and eyelashes. Thus, the correct recognition rate decreased in the existing approaches due to occluded eyelids and eyelashes classified as an iris region. Therefore, this research proposed an iris image model and a feature extraction method for partial iris recognition. Several issues that are investigated in this research i.e. which parts or regions of the iris are more stable and provide distictive texture patterns, what is the optimum coverage area of the iris that is required to extract discriminant texture features for partial iris recognition, the existing feature extraction methods and the nature of features whether they can be used for partial iris recognition