关键词: 指纹识别;ARM; Linux;移植;预处理算法 Abstract Fingerprint recognition is one of the most widely applied biometrics at present. As much more attention is paid to security, fingerprint recognition technology is applied in more and more domains such as authentication's method and so on. Benefit from the semiconductor technology development, the automated fingerprint identification system cost is reduced rapidly, which promotes further application of fingerprint recognition technology in real world. First, fingerprint image preprocess was studied, including segmentation of fingerprint image background, normalization, image enhacement, binarization based on dynamic threshold and thinning processing. After image preprocess, fingerprint feature can be extracted from the ‘skeleton’ of fingerprint image, and fingerprint matching algorithm was studied as well. On the considerataion of the market requirements for fingerprint recognition system, a CMOS optical fingerprint gathering identification system, with low costs and high performance, is designed based on the ARM&Linux system platform. On the basis of advantages and disadvantages analys