Wavelet Transform And Reverse Determining Technology Of ABO Blood Group

Over the past several decades, Wavelet analysis has been quickly developed asa modern subject because of its excellent characteristic of localizing both in timeand in frequency field. Wavelet analysis is very significant in fundamental theoryand factual application. At present, wavelet analysis has drawn much attention forresearchers in the field of image processing, which is an advanced method toprocessing medical image. And the medical image processing plays an importantrole in clinic diagnosis and therapy, which may increase the accuracy of thediagnosis and economize time for the doctor.How to analyze image of the determination of reverse ABO blood group isone of the important studies for medical image processing and analysis. In thisthesis, wavelet analysis is used to analyze the sample image of the determination ofreverse ABO blood group. The development of wavelet analysis will be improvedby factual application in the determination of reverse ABO blood group and others.Thereinafter, a few parts of content are researched in this thesis:The first section, the theory of wavelet, de-noising modified method and edgedetection based on wavelet transform in the determination of reverse ABO bloodgroup are introduced.The second section emphasizes the theory of wavelet transform. Wavelettransform is a kind of local transform in time and frequency field. And it mayeffectively distill from the signal. It is entitled"the microscope of the Math"by thescholars.The third section focuses on the applications, component of the check and testof the determination of the reverse ABO blood group. The color of the sampleimage has no effect on the result of image processing. We will emphasize thede-noising and edge detection of grey image.The forth section mainly illustrates the de-noising methods based on wavelettransform. Donoho's threshold method is most widely applied. Due to its singlethreshold value, however, it cannot separate signal from noise completely at eachstep of scale and hence its de-noising effect is not satisfactory. By comparativelyanalyzing, Donoho's hard threshold and soft threshold method, this paper putsforwards a new method. The simulation de-noising result shows that betterde-noising effect can be achieved by this modified wavelet, which combines theadvantages of both methods. And the signal to noise ratio can be improved.In the fifth section, the operator of the traditional edge detection is brieflyintroduced. The paper describes wavelet transform and the principle of itsapplication in edge detection. It is further pointed out that the idea is wellapplicable in the fast recognition of targets.Finally, the last section is summarization and expectation.