Design of fingerprint recognition access control system based on ARM9

Fingerprint access control system is a high-tech security facility based on biometric recognition technology. In recent years, it has been widely used at home and abroad and has become one of the symbols of modern building intelligence. For some core confidential departments, such as important agencies, scientific research laboratories, archives, civil aviation airports and other places, fingerprint access control system can provide efficient, intelligent and convenient authorization control. Because fingerprints are easy to carry, different for everyone, and unchanged throughout life, using fingerprint recognition as a means of identity authentication greatly improves security and credibility compared with traditional keys and passwords.

The system is based on the ARM9 chip Samsung S3C2440AL, Veridicom's fingerprint acquisition chip FPS200 as the hardware platform, and embedded Linux as the software platform. In this research field, the recognition system based on the PC platform has always been the focus of research. The system based on the ARM platform implemented in this paper has the advantages of being light, easy to install, and low cost, and has good development prospects.

1 System Hardware Design

The main frequency of S3C2440AL is 400 MHz, and the highest is 533 MHz; the FPS200 fingerprint sensor consists of a 256×300 capacitive sensor array, with a resolution of up to 500 dpi, an operating voltage range of 3.3 to 5 V, an 8-bit ADC inside the sensor, and 2 sets of sample-and-hold circuits. The block diagram of the entire system is shown in Figure 1.

2 Operating System

Since embedded Linux has the advantages of small kernel, high efficiency, open source, and many platform tools, the system uses embedded Linux as the operating system platform. The main steps to build the platform are as follows:

(1) Download U-boot via JTAG;

(2) Configure the Linux Kernel and download it through the serial port;

(3) Develop the FPS200 driver and load it dynamically.

3 Fingerprint recognition algorithm flow

According to the main functions in the identification process, the fingerprint recognition system can be divided into fingerprint image acquisition algorithm, image preprocessing algorithm, feature extraction algorithm, and feature matching algorithm.

(1) The fingerprint image acquisition algorithm flow is shown in Figure 2.

(2) The fingerprint image preprocessing algorithm flow is shown in Figure 3.

(3) Fingerprint feature extraction algorithm process

Before extraction, it is necessary to first remove the pseudo-fingerprint feature points and then extract the topological data structure of the fingerprint.

(4) The fingerprint feature matching algorithm flow is shown in Figure 4.

4 Fingerprint Segmentation Algorithm Principle

Suppose the number of pixels of a fingerprint image is N, it has L gray levels (0, 1, 2, ..., L-1), and the number of pixels with gray level i is ni, then the image histogram is normalized and has a probability density distribution:

Assume that the threshold t divides the image into two categories, C0 and C1 (i.e., object and background), where C0 and C1 correspond to pixels with gray levels {0, 1, 2, ..., t) and {t+1, t+2, ..., L-1} respectively. The occurrence probabilities of C0 and C1 are:

From Figure 5 we can see the effect diagram obtained by using this method, and the segmentation result is good.

5 Conclusion

This paper uses the ARM9 processor as a platform to systematically implement an embedded access control system with good recognition results. The entire algorithm flow of the system design is given, and the image segmentation algorithm is introduced in detail. Compared with previous research results, the system has the advantages of simple platform, high recognition rate and fast recognition. However, the system's processing effect on distorted images is not very ideal, and needs to be further strengthened in future research.