Adaptive Wavelet Threshold Method For Fingerprint De-noising Based On Wavelet Analysis

With the development of science and technology, identity recognition technology is more and more widely applied in people’s daily life. Fingerprint feature is unique to each persons, it can be used as the basis of identification. Wavelet theory is extensively used in the field of signal processing because of its rapid development and good time-frequency characteristics.This paper discusses the application of wavelet threshold de-noising method in fingerprint image de-noising. An adaptive de-noising method based on soft and hard threshold function is proposed. The new way can deal with the noise of different intensity in a certain range adaptively.The first chapter introduces the background and significance of threshold de-noising method and the structure of this article.In chapter 2, we give introduce the basic knowledge of wavelet transform. It states the principle of Visu Shrik de-noising method and lists several different ways of threshold quantization.In chapter 3, we analyze the advantages and disadvantages of soft, hard threshold function and some improved threshold functions. For convenience, we call the improved function for compromising function. By theoretical analysis, on the one hand, the attenuation degree of wavelet coefficients that controlled by compromising threshold parameter is slow. On the other hand, the wavelet coefficients do not converge to itself. Based on the consideration of the above two points, an improved compromising threshold function is proposed, then we analysis the new threshold function from the perspective of theory. Then, we investigate the risk of the improved compromising threshold function. The Experimental results show that the improved compromising threshold function de-noising method has better results than the soft and hard threshold function de-noising method, and the parameter in the first layer has a relatively large impact on the de-noising effect.In chapter 4, according to the experimental data, we establish three groups adaptive parameter control model with dichotomy. Simulation results show that within a certain range of noise intensity, Gaussian parameter model has better de-noising performance.We compared the adaptive method with several advanced methods at home and abroad in the fifth chapter, the results show that the adaptive method has better de-noising effect than other methods, and with the increase of noise intensity, it has higher de-noising efficiency.In chapter 6, we summarize this article and present some unresolved issues and the future research directions.