Research On Fingerprint Image Enhancement Based On DWT

With the advent of the network society and the growing popularity of e-commerce, traditional system security technology which depends on token or password has been facing serious challenges.Attendant biometric identification technology has become a hot research.Automatic Fingerprint Identification Technology is the most acceptable,automatic and mature biometric identification technology. However, in practical application,a large part of fingerprint images are of poor quality, in order to extract reliable fingerprint characteristics and to provide reliable basis for fingerprint identification,and even to improve the performance of the whole system,it needs an effective enhancement processing.This paper just studied and discussed the fingerprint image enhancement.The wider use of wavelet transform in recent years provides fingerprint image processing with a new research method.As a new mathematical tool, due to its multi-scale, multi-resolution of the inherent characteristics , wavelet transform shows the following advantages in the field of image processing:the perfect reconstruction ability of wavelet transform ensures that the signal is no loss of information in its decomposition process,and it does not produce any redundant information;The image is decomposed into approximation part and detailed parts which represent different structures of the image by wavelet transform. Therefore, the structure information and details of the original image can easily be extracted; two victoria wavelet decomposition provides the image analysis with the direction selectivity.Based on the above analysis,this paper studied fingerprint image enhancement based on wavelet transform. The main contributions are summarized as follows:1. Proposed an integrated denoising algorithm based on mid-value filtering and wavelet transform of fingerprint image.According to the existed type of random noise in fingerprint image,we proposed an integrated denoising algorithm based on mid-value filtering and wavelet transform of fingerprint image. We researched deeply in the application of wavelet transform in fingerprint image denoising,including:the choice of wavelet fundarment, the choice of decomposition level, the ralation between the wavelet transform modulus maxima and Lipchitz index, the analysis of wavelet coefficients with noise in fingerprint image, wavelet threshold selection, etc. On the basis of the study,we designed and applicated the denoising algorithm based on the wavelet transform modulus maxima and wavelet threshold.Fingerprint images provided by FVC2004 were denoised with the algorithm we proposed on Matlab 7.0 platform ,and achieved better results. 2. Proposed an enhancement algorithm based on wavelet transform of fingerprint image.Accordingly there is often not clear enough ridge structure, breakpoints, even forks and other structural noises in fingerprint image,we proposed an enhancement algorithm based on wavelet transform of fingerprint image. On the basis of previous studies,using the wavelet transform characteristic of high-low-frequency decomposition,the algorithm first got the accurate direction and frequency information of the ridge in the low-frequency domain, not only to avoid the interference of high frequency noise signal to the direction and frequency information of the ridge,but also to save time. Then according to the different characteristics of each sub-band after wavelet decomposition and the different enhancement objectives of them, different filtering methods are used.We combined the advantages of Gabor filter and the direction filter appropriately. Finally wavelet reconstruction was proceeded. Experiment shows that the algorithm has been made better enhancement effect and higher time efficiency.on the basis of previous studies,this paper studied and discussed the fingerprint image enhancement based on wavelet transform.I hope it can play a certain role in promoting further research work in the relevant areas.