Research And Design Of The Key Technologies On Automatic Fingerprint Identification System

With the evolvement of information technology and impelling of information revolution, the domesticity of people become opening and the requirement about safe and convenient information communication is on the increase. The traditional identity authentication technique does not satisfy the need of development in smart home gradually. As compare with the traditional technique, the automatic fingerprint identification technique has predominance because of one's unique, convenient and highly reliable characteristic. It is provided an appropriate solution.In this dissertation, fingerprint identification technique is studied and the whole fingerprint identification system frame is constructed in the information home environment. Based on the discussion and analysis about the key techniques, the design of system is accomplished and the all-sided evaluation of this design is given. Especially, tow kinds of important problem is solved thereinto.The fingerprint acquisition which has good quality and high speed is the precondition and the assurance of AFIS. In the article, the USB and fingerprint identification technique is integrated and acquisition of fingerprint data is implemented. As the traditional acquisition technique hasn't good adaptability, acquisition feedback technique is adopted and the fingerprint acquisition is controlled in real time. The acquired fingerprint image is adjusted automatically and optimization is implemented.A complete set of pretreatment algorithms of fingerprint image is designed and implemented. Firstly, the error of traditional directional image is corrected and fingerprint image is enhanced by the directional filter. Secondly, the misjudgment of the adaptive threshold is amended and the exact binarization is realized. Lastly, the common thinning algorithms are improved and two-step thinning algorithm is proposed. This thinning algorithm has good performance, so the post-processing is simplified.