面向嵌入式平台的虹膜身份确认技术-控制科学与工程专业毕业论文.docxVIP

面向嵌入式平台的虹膜身份认证技术 面向嵌入式平台的虹膜身份认证技术 IRIS IDENTIFICATION TECHNOLOGY FOR EMBEDDED PLATFORM ABSTRACT Iris recognition is an identification technology based on the human physiological characteristics. It has the advantage of uniqueness, stability, high recognition rate and non-invasive. With the development of embedded systems as well as the maturity of iris recognition technology, iris recognition applied to embedded systems becomes a research hotspot. The research on iris recognition algorithm running in embedded system with high reliability and high instantaneity will have very high practical value. The paper is studied based on the iris recognition algorithms that have high recognition rate. Iris recognition system consists of iris image acquiring, image preprocessing, feature extraction and matching. This article firstly studied the existing iris localization, feature extraction and matching algorithms. Then make some improvements focus on the characteristics of embedded systems. In the process of iris localization, the paper firstly use the method of gray binarization to locate the edge of the pupil, and then use the Canny operator and the improved Hough transform method to locate the edge of the iris edge. The paper mainly using methods based on line extraction process to normalized iris. It has only one analysis model and is easier to understand. In the feature extraction, the normalized image is divided into several parts first, and then analysis image texture feature by wavelet filter. After that, the paper use reciprocal variance weighted Euclidean distance classifier to classify the image. Finally, the whole process of iris image recognition is realized by Qt under Linux system programming. This thesis’ experiments are based on the Chinese Academy of Science Institute of Automation eye image database. It has proved that the improved algorithm has a high recognition rate with a small footprint, fast running speed advantages and is suitable for embedded systems.