Application Study Of Fingerprint Identification Based On Linux Embedded System

With the continuous development of science and technology, the security of personal information have higher requirements for people. Biometric technology is therefore increasingly be of great importance to researchers. Among the biological identification technologies, fingerprint recognition technology for its easy-to-use, high accuracy, low-cost advantages such as a matter of considerable concern in the criminal investigation, e-commerce, information technology has broad application prospects.Embedded systems will not only make the speed getting faster and faster, but also richer interfaces, which is more suitable for a variety of electronic products developing. The combination between fingerprint identification technology and embedded systems allows the combination of fingerprint identification system break away from the space and environmental constraints, more widely applicable to the needs of a variety of occasions. In this paper, a preliminary study is done on embedded fingerprint identification system. The following three areas are completed:1. The structure of fingerprint identification system is designed based on the fingerprint technology status and development direction at home and abroad. PXA270 processor, FPS200-type fingerprint sensor and embedded Linux operating system are used to build fingerprint identification system.2. The software platform of the embedded fingerprint identification system is designed. The device driver implementation principle of Linux operating system is analyzed. FPS200-type fingerprint sensor driver is compiled into the kernel. The cutting and transplantation of the kernel are investigated and the system kernel is compiled.3. The fingerprint image identification algorithm is studied and the corresponding experimental verification is obtained. The algorithm includ image segmentation, image enhancement, binarization, thinning, feature extraction, the removal of pseudo characteristics. The experimental results show that the above algorithm is suitable for fingerprint image recognition.4. The software graphical interface of the embedded fingerprint identification is designed. A graphical interface is designed using of Qt Designer tool in the host Linux operating system , then a program is generated which Qt / Embedded and Qtopia can run. At last, installation and implementation of the program is completed.After experimental verification, the designed development platform can run a Linux system, transport the fingerprint data and display graphical interface.