基于ARM9的指纹识别门禁系统

为了实现基于ARM的指纹识别门禁系统,采用Veridicom公司的FPS200指纹采集芯片进行指纹采集,采用Samsung公司ARM9 S3C2440AL给出了系统的软硬件设计及仿真结果。经验证该系统拒识率小于0.1%,认假率小于0.01%,识别时间小于2 s,实验结果良好。此外,重点介绍了该系统中采用的指纹分割算法,该算法以前景与背景类间方差最大为原则,分割稳定的同时具有分割阈值的自适应性。

Access System of Fingerprint Recognition Based on ARM9

The hardware software design and the simulation result are proposed to realize the ARM-based access system of fingerprint recognition. Fingerprint sensor FPS200 and main processor S3C2440AL are adopted for the design. The conclusion of high veracity （FAR〈0.01%）, low rejection rate （FRR〈0. 1%） and short recognition time （〈2s） can be drawn by the simulation. Additionally, the algorithm of fingerprint segmentation with principle of maximum inter-class variance is described in detail, the self-adaptive threshold can be obtained with the stable segmentation result.