The Research And Implementation Of Iris Recognition Alogorithm Based On Qt/Embedded

With the increasing requirements for security, biometrics based on personal identification methods have received extensive attention. Recently, iris recognition is becoming an active topic in biometrics due to its high reliability for personal identification. A great deal of progress in iris recognition has been achieved in the past decade, but many efforts remain to be taken to further improve the performance of iris recogniton systems.According to different platform there are two kinds of system composing for iris recognition system: one is based on PC and another is based on embedded system. The embedded iris recognition system has become the main development directions of iris recognition system with the requirements of miniaturization, high reliability and low power consumption for iris recognition system. With the deepening of the application and improvement of industrialization for application of embedded iris recognition system, the system makes more stricter demand of software part such as high reliability, instantaneity, adaptability. Thus, the research on iris recognition algorithm running in embedded system with high reliability and high instantaneity, has very high practical value and realistic significance.The iris recognition algorithm which has high instantaneity and short code length on the premise of high recognition rate is studied, the algorithm can well support embedded system, we put forward our improved algorithm after investigating some of the key issues related to iris location and iris recognition, and the algorithm is realized with Graphical Development Tools Resource Qt/Embedded which is based on C++ using linux operating system as developing platform. The main contents and innovative contributions of this dissertation are as follows:(1) This thesis put forward a fast iris location algorithm. In this thesis, two steps of coarse location and exact location is used to locate the interior boundary of iris, the algorithm avoid the blindness of searching. The algorithm first make use of the geometrical features and gray level distribution features of pupil to locate the center of the pupil roughly, and then use an improved algorithm based on active contour to get the center and radius of the pupil accurately, we use an improved circle detection operator algorithm which search the center and radius of the outer boundary of iris within small area, many algorithm assume that the center of interior boundary and center of outer boundary are the same, but in this paper, the algorithm improves the accuracy and enhances the adaptability of iris location by setting different location of the two centers of iris in the context of not increasing much algorithm complexity. The experimental result shows that the algorithm can reduce locating time without prejudice to locating result and it can meet the requirements of accuracy and instantaneity for embedded system.(2) In this chapter we put forward a reliable iris recognition algorithm with low memory space and high speed. According to the characteristic of iris texture which is sparse outside and rich inside, iris is divided into two bands and every band is filtered by the 2D-Gabor filters and encoded with the phase of the result of filtering. Because there are less texture information in outboard band than that in inboard band, multiple Gabor feature in the same scale with different orientations are fused by using the magnitude information, and the iris codes are generated based on the phase information when we deal with the outboard band. Compared with traditional non-fusion approach, the proposed approach has the same high recognition performance, but the size of iris codes in the proposed approach is only three quarters of the traditional one, so this approach can support embedded system well with its high speed and low memory space.(3) After analysing the frame structure of the iris recognition application program based on Qt/Embedded development environment, the GUI(Graphical User Interface) and experimental result of iris recognition system are given under the embedded development environment that we set up.