Research On Image Enhancement Based On Dyadic Wavelet Transform

Medical image is a kind of special digital image recording for certain parts of the human body or the focus position information,which is reflected on the path of rays to penetrate the tissue of the human body physiological X ray absorption amount of the accumulated value,to its application throughout the clinical medical work.The image acquisition and transfer process due to the image of its own mechanism,influenced by the environment,ex-ternal noise and other factors,there is overall brightness dim,low contrast features.This has seriously affected the analysis and understanding of the medical image.Therefore,a good medical image enhancement method can not only improve the map clarity at the same time,but also can reduce the fuzzy image enhancement is the main purpose of the image to improve the polluted or damaged in the preservation of image detailsHowever,the remote sensing image is a special digital image in keeping an account of features information,with a large amount of information,the overall darker and the target and background can not distinguish characteristics of so how to maintain a balance between the image details and noise are scholars in different countries continue to strive to the di-rection of the main.The purpose is to improve the contaminated or damaged image,to highlight the characteristics of the image itself.While in the process of how to avoid noise amplification and taking into account the texture and edge enhanced image detail,is still a problem to be solvedBased on the existing methods of medical image enhancement,a thorough study of the basic theory of dyadic wavelet transform and its lifting scheme,and the main properties and applications of Contourlet transform,multi-scale Retinex and fuzzy theory.The main work and structure of the article are as follows:In the first place,the background and significance of the research and the current re-search status of image enhancement at home and abroad are introducedIn the next place,the main classification of spatial enhancement is described,and the anti-sharpening mask and fuzzy enhancement algorithm are described in detailAdditionally,we introduce Fourier transform,wavelet transform,dyadic wavelet trans-form and its construction based on dyadic wavelet filter and Contourlet transform respec-tively.According to its algorithmin characteristics,advantages and disadvantages image en-hancementProposed the DyWT and fuzzy set theory for image enhancement method,based on initial B-spline wavelet and lifting dyadic wavelet,the low frequency coefficients after de-composition using multi-scale improved Retinex algorithm to adjust and improve the image brightness uniformity distribution.According to the total body of high frequency coefficient obeys the generalized Gauss distribution,using adjustable factor Bayes atrophy the thresh-old denoising method of high frequency coefficient denoising processing,and through fuzzy contrast enhancement to further improve the global image contrastProposed (?) Trous algorithm of remote sensing image enhancement based on fuzzy set.This method of remote sensing image enhancement can significantly improve the image contrast in the subjective visual effect based on has significantly improved,the objectively image mean and signal-to-noise ratio have greatly improvedIn the end,a new method of image enhancement is attempted,which combines the multi-scale Retinex,Contourlet transformation and improves the contrast of the blurred image,and proposes a Contourlet domain fuzzy contrast image enhancement based on multi-scale Retinex.It is shown that this method enhances subjectively the enhancement of the image contrast and the improvement of the visual effect,and objectively the mean value of the image and the signal to noise ratio are greatly improved.