Iris Image Recognition Based On Support Vector Machine

Iris recognition has received much attention in the present. The research and application of iris recognition has become a hot research direction, which makes iris recognition technology to be popularized and applied quickly. Iris recognition technology has been used in many fields, such as access control system. The research of iris recognition system has a broad and far-reaching significance.A new method for iris feature extraction is proposed in this paper. Because the iris image is affected by the upper eyelid and eyelashes, this paper tries to extract the full iris, which is affected by the upper eyelid and the lower part of the eye.First, using the gray histogra realized the accurate location of the iris inner boundary; subsequently, to overcome eyelash and eyelid occlusion gives a kind of iris boundary localization method to obtain the accurate iris effective part. Finally,by using normalization algorithm, into by annular iris feature for rectangle features for the training and recognition of the iris. Through experiments, the method is accurate and effective.After obtaining more accurate and effective iris features, support vector machine(SVM) is adopted for the recognition and classification of iris images.