| Touchless fingerprint recognition technology has become a hot spot of research in the field biometric technology in recent years, and has a wealth of technical reserves and broad application prospects. It overcomes the traditional fingerprint recognition technology fingerprint defects in the acquisition mode, with the advantages of easy to collect and accept,which made it a hotspot biometric research. With the rapid increase of the smartphone market and the rapid development of the performance, the application of touchless fingerprint identification technology has opened up a field of more practical value. With the development of large-scale popularization of mobile Internet technology, the emergence of mobile payment methods, which are urgent need for a more convenient and effective authentication techniques to ensure that the phone user's mobile payment security and information security,which also coincide the development of touchless fingerprint recognition technology.This paper analyzes the existing principle of touchless fingerprint recognition system,including fingerprint image preprocessing. feature extraction and matching, based on this research of- touchless fingerprint recognition technology study for smart phone, and study the key technologies Research. Then select the most widely used operating system, the Android platform, combined with the open source computer vision library OpenCV, then develope a touchless fingerprint identification system based on the Android platform.First, study the the research of traditional fingerprint recognition algorithm, select the appropriate touchless fingerprint image pre-processing and image enhancement algorithms for the subject. First the YCbCr space are used to extract fingerprint areas, then compensate the extracted fingerprint image illumination and filtering de-noising, and the extracted image are transformed from RGB to gray and normalized. Using the homomorphic filtering method to enhance the the contrast of the fingerprint ridges and valleys, and then proposed an improved Log-Gabor filter method, which can enhance the fingerprint image to improve the clarity of the fingerprint image, these methods are proposed for the characteristics of touchless fingerprint image. The results show that the algorithm used in this paper not only accelerate the processing speed, and has the the reasonable computation and storage calculations, but also has better robustness and recognition rate.Secondly, study the principle of fingerprint recognition algorithm, we select a smaller calculation algorithm analysis for consideration of the computing power of the phone, in order to pave the way of the transplant Android system. Based on ELDP algorithm, an improved algorithm named MK-ELDP is proposed. Firstly, the method divides the fingerprint image into several blocks, and replaces the block gray value of this area. Then, the ELDP features of the mean images can be calculated. Lastly,the recognition result can be obtained using histogram matching method. Through the verification of experiment, comparison with other feature extraction algorithm,the proposed algorithm not only reduces the computational complexity, improve the recognition rate, but also has a good recognition rate of the fingerprint image.Finally, combined with the preceding analysis of touchless fingerprint image algorithm,these algorithms will be ported to the Android platform. This article design touchless fingerprint recognition system on the Android platform, and on the Android platform we complete the function of the acquisition of the fingerprint image, preprocessing and fingerprint recognition. Based on the operating system Android4.2.2, we tested multiple fingerprint images to verify the feasibility and effectiveness of the method,in order to lay the foundation of further research for the touchless fingerprint identification technology,for personal authentication and other related applications, at the same time show a good practical application prospect and practical value. |
PDF Full Text Request