Fingerprint Core Point Location Algorithm Based On Sliding Window

With the rapid development of technology, security issues increasingly become a cause of concern. As one of the main technical information systems to achieve access control, the authentication technology has been widely used and valued. The traditional password-based identification technology and smart card security exist security issues, practical problems etc to be resolved, so that the authentication technology has become a focus of current research in the field of information security. Because of unique and security etc, biometric technology becomes the research hotspot in the identification direction.Because of its ease of acquisition, less amount of data processing, etc, fingerprint has become an important biometric technology branch and has been widely used. Fingerprint image edge is the most obvious changes in local gray area of the whole image, and the contour information includes fingerprint geometry which is an important basis for the processing of image features in the process of fingerprint recognition and image segmentation process. Fingerprint center is a key feature of fingerprint texture. The fingerprint image edge detection pre-processing and center location are important parts of the fingerprint identification whose results will directly affect the effectiveness and efficiency of fingerprint matching. Thus, the fingerprint center location technology has important significance of theory and implementation.This paper focuses on problems of fingerprint edge detection and center location methods and studies improved methods, and designs two kinds of edge detection and center location methods based on fingerprint data. The main work shows as follows:1. Fingerprint edge detection algorithms based on Hilbert-Huang transform (HHT). Compared with the existing Sobel operator, Laplace operator, Canny operator, the wavelet algorithm on the detection of the fingerprint edge, HHT transform shows the great advantages include good detectability etc. According to the IMF components which come from the EMD decomposition in HHT transform, whose distribution characteristics are from high frequency to low frequency, the alone IMFl component and the combination of the IMF1 and IMF2 are conducted edge detection respectively. The results show that the fingerprint image edge detection results of the latter is clearer, and retains more fingerprint detail.2. A method of the fingerprint core location algorithm based on sliding window. The main idea is to combine sliding window with core location algorithm of complex filter. First of all, according to the principle of statistics, determine moving path of the sliding window and performance indicators. Then detect core location of fingerprint information within the sliding window by using the complex filter core location algorithm. The results show that on the premise of the positioning accuracy, the testing time of the fingerprint core location algorithm based on sliding window is greatly shortened which improves the efficiency of the algorithm.