An Iris Processing using the Characteristic Value
Lunman Deng, Sek-Koun Youn, Kyu-Boem Lee, Jeong-Young Song · 한국정보기술학회논문지 · 2011
In general, it is reported that biometrics recognition rate is low if it is using iris information. Iris Information Locator inject expensive and difficult to recognize the difficulties caused by non-contact. Initial located point in pupil is gotten by average gray method. In order to fasten the speed of iris location, this paper introduces a fast location algorithm for pupil detecting based on the geometrical characteristic of circle. The use of iris feature extraction independent component analysis and core vector machines and geometric information of identification. According to the characteristic of ICA classification, the SUM and CUM after identification method, recognition rate is 90%-91%. The experimental results show that this method can be progressive.