The Purpose Of Ear Recognition Based On Force Field Transformation

As a relatively new kind of biometrics, the purpose of ear recognition is determining or validating people's identity based on ear images. Ear recognition can be a beneficial supplement for other biologic recognition or be solely used in certain situation. Since ear recognition is in the primary research, there is not sufficient theory about it. Ear recognition needs to refer to other biological recognition technology and theory. And there should be a particular recognition method because of the special physical structure and position of ears.This thesis focuses on the key steps of ear recognition such as universal gravitation field transformation and feature extraction. It tries to enrich the method of ear identification and recognition. It would be my pleasure that the paper can be beneficial to other researchers.As we know, the final object in determining feature space is to reduce the dimensionality of the original pattern space,while maintaining discrimination for classification. To meet the requirement of ear biometrics, a new universal gravitation field transformation is put forward to treat the image as an array of mutually attracting particles that act as the source of a universal gravitation field. Accompany with the force field there is a scalar potential energy field,which takes the form of a smooth surface that resembles a small mountain with a number of peaks connected by ridges. The peaks correspond to potential energy wells,while the ridges correspond to potential energy channels. This paper describes how to extract force field features using an universal gravitation field transformation algorithm, and exploit the field lines to automatically locate these channels and wells,which form the basis of ear features. It is shown that ear recognition seems to be a potential biometrics,and the force field transformation method is especially robust to noise. One of the most distinct advantages of our method is that the contour of the auricle does not need to be explicitly extracted from the background.In conclusion, the key steps of force field transformation are studied and a good experimental result is obtained,which is expected to promote the development of earrecognition in a certain extent.