Personal Recognition With Ear Biometrics

Research of ear recognition technology, as well as its application, is a new subject in the filed of biometrics recognition. Auricle has stable and rich structures. It can't be affected by countenance and it is difficult to be cheated, easier to be collected, more acceptable. Therefore, ear biometrics is becoming a new hotspot of biometrics. Now ear recognition just gets of the mark, and its feature extraction method and recognition rate is not satisfactory.In this thesis, the basic concepts of ear recognition are introduced, and the existing methods of ear recognition are surveyed. The research work is done in two stage, contour extraction and location, feature extraction and classification.According to these characteristics, this paper introduces a force—field Fisher classifi er for ear recognition. This method, which is robust to changes in illumination, applies the Fisher linear discriminate analysis to an augmented force-field feature vector derived from the force—field transformation of ear images. The feasibility of the new method has been successfully tested for ear recognition. This method even achieves 98 % recognition accuracy for ear images from selected database.This thesis introduce a optimize method during feature extraction and classification. Improving arithmetic speed and lessening the reverse in the process of conversion course make use of fourier transform and reverse transform, and give out the procedure source code. At last, applying the extracting characteristic of force field and the Fisher linear discriminate to make out the recognition accuracy.The experiment results demonstrate that the improved force field ear recognition has greatly improved the accuracy of the feature extraction and classification, and becoming the basement for the future research and application of ear recognition.