Fingerprint Identification

The characters of personal fingerprint have the proper advantages in the biological identification field, in which everyone has his own unique fingerprint which can identify himself. Because of the more convenience and security in the collection of fingerprint, there are more advantages in the reliability of identification than other biological identification technology. So, there is more application prosperity in the future. The processes of the fingerprint mainly comprise the collection of fingerprint data, preprocess, extracting of minutiae, matching and the results. Because the preceding outputs are just the following inputs, the effects of the every process will directly affect the following process' precision. This paper bats around the algorithms of the all processes in the auto fingerprint identification system and then improve the current algorithms relevantly so as to find out the algorithms adapt to the fingerprint and improve the capability of the identification system. This paper discusses and improves the key algorithms of auto fingerprint identification system. There are 7 chapters in this paper. The first chapter is introduction, in which mainly introduce the current development status in quo of the fingerprint identification technology and the processes adopted abroad. The seventh chapter is the conclusion. The other five chapters respectively research the following aspects. Research the extracting of fingerprint direction. From the experiments, realize the RAO algorithm and the rapid algorithm from the grey picture directly of fingerprint. Then analyze and compare the effects of the extracting of the two methods. The compare chart of extracting direction of fingerprint shows clearly that the direction based on grey algorithm around the singularity points vary acutely. We can have the hint from the rudimental area of the singularity points' locations through the un-direction degree that reduce the target range without the computation to the all picture every time. As the result, the time-consuming will be reduced in the 3 chapter and because of the more accurate precision than RAO algorithm; the iteration of extracting will also be reduced. So, this algorithm can be still improved by the deeper analysis. By computing the direction of fingerprint based on the grey by which is used to locate the singularity points in 3 chapter and reduce the target range based on the direction information, the time-consuming will consumedly be reduced which both improve the precision of the whole auto identification system and improve the speed of the algorithm. Locate precisely the core points and delta points of fingerprint so as to confirm datum marks from the singularity points set in matching algorithm. Accelerate the matching and improve the matching algorithm based on the curve fitting. The purpose of locating precisely the core points and delta points is to improve the point-matching algorithm based on the curve fitting. First, select datum mark from the singularity set. From the conclusion of matching analysis in 6 chapters, if the fingerprint is an arch, we'll use curve fitting algorithm to confirm the datum mark from the minutiae's set. This improvement will both accelerate the finding of datum mark in matching algorithm and mend the problem in original location of the datum mark. At the same time, the improved algorithm depends on the precise location of singularity. So, the precise location of singularity is one of the most steps in the whole identification process. In this chapter, iterate to compute the direction of fingerprint so as to improve the precision. Though there are more time-consuming than before, it's worth. Precise location of singularity points will simplify the algorithm of matching and improve the precision. Different algorithm of direction will also improve the performance of this algorithm. Realize two kinds of two-value algorithms and two kinds of thin algorithms, which are the OTSU method based on the whole picture, self-adaptation method, Hilditch thin algor...