Font Size: a A A

Rotation Invariant Iris Recognition Method Based On Non-separable Wavelet

Posted on:2010-02-03Degree:DoctorType:Dissertation
Country:ChinaCandidate:J HuangFull Text:PDF
GTID:1118360302471102Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Biometric recognition refers to the automatic recognition of individuals based on physiological or behavioral characteristics. Physiological characteristics,such as face,iris, finger-print, palmprint, voice etc. are connatural, whereas behavioral characteristics which are the habits of human, such as gait, handwriting, are postnatal. Compared with traditional personal recognition methods, biometrics has the following merits: (1) need not to remember intentionally; (2) have little risk to be forged; (3) available anytime anywhere.Iris recognition is performed mainly based on the intricate structure with many minute characteristics such as furrows, freckles, crypts, and coronas etc. Iris recognition is proved to be one of the most reliable biometrics in terms of identification and verification performance.In this dissertation,some of the key issues related to iris recognition system are investigated under the consideration of the iris physiological structures. The main contributions of this work are as follows:(1) We first introduce the non-separable wavelet transform into iris recognition. Compared with the separable wavelet transform, its high frequency components can reveal more singularities reflecting various orientations of the image than the separable one can do.(2) A self-adaptive method of iris preprocessing is proposed in this paper. No matter what shapes the iris boundaries are, the proposed method can segment the iris area accurately. The conventional methods of iris preprocessing are based on the assumption that both the inner boundary and the outer boundary of an iris can be taken as circles, and segment the iris using the two circular borders. However, we investigate the iris boundaries in the public iris databases, and find that the actual iris boundaries are not always circular. In order to solve this problem, a new self-adaptive approach for iris preprocessing is proposed. It can detect the iris boundaries adaptively regardless of the shapes of the boundaries. (3) Two rotation invariant methods of iris feature extraction and matching are proposed in this paper. In the recognition system, it is desirable to obtain an iris representation invariant to translation, scale, and rotation. In the conventional algorithms, translation invariance and scale invariance are achieved by normalizing the original image at the preprocessing step. But these algorithms cannot achieve true rotation invariance. They only achieve approximate rotation invariance at the matching step. Some of them achieve approximate rotation invariance either by rotating the feature vector before matching, or by registering the input image with the model before feature extraction. In the other methods, there is no explicit relation between features and the original image. They thus obtain approximate rotation invariance by unwrapping the iris ring at different initial angles. They store several additional iris codes for different initial angles. During matching, the test iris code is compared against all stored ones and the minimum distance is chosen. To overcome the above disadvantages, two new rotation invariant approaches for iris feature extraction and matching are proposed. Different form the conventional algorithms, the proposed algorithms achieve rotation invariance at feature extraction step. The iris features are rotation invariant. So, in our methods, the complexity of the methods are not increased, and the error rates of the methods may never increase, no matter how much the iris rotate. Experimental results on public iris databases show that the FAR and FRR of the proposed approaches are desirable.
Keywords/Search Tags:Biometrics, Iris Recognition, Iris Segmentation, Normalization, Feature Extraction, Matching, Non-separable Wavelet
PDF Full Text Request
Related items