Study On Algorithm Of Image Processing For Medical X-ray Gastroenteric Digital Image

X-ray medicine imaging has become essential and necessarily tools for the doctors in modern iatrology times. Quality of x-ray medicine imaging would directly affect doctor's diagnoses and cure to patients. However, during the process of obtaining x-ray medicine image, the imaging will be inevitably intermixed by various of noise. High resolution x-ray gastrointestinal digital image is the same, too. While normal organized x-ray attenuation modulus are very close to the Gastrointestinal pathological changes modulus and along with noise from the image, the doctor will have no way to diagnose based on patients'gastrointestinal characteristics. Therefore, it is precondition for Doctors'accurate diagnosis that you have to master excellent high resolution x-ray gastrointestinal digital image. It can be seen that to process the noise of high resolution x-ray gastrointestinal digital image and try to reduce noise blight of high resolution x-ray gastrointestinal digital image will be one of effective way to improve the quality of high resolution x-ray gastrointestinal digital image.The thesis based on describing medical x-ray imaging system structure and principle, according to imaging process of high resolution x-ray gastrointestinal digital imaging system to research and analyse various noise resources and noise characteristics, and then point out that the noise can be divided into systemic inherent characteristics noise and systemic random distribution noise. By optimizing hardware configuration, it can reduce systemic inherent characteristics noise; via analysis of systemic random distribution noise, it mainly includes Salt & Pepper noise and Gaussian noise. Up to now, after analyzed and compared traditional digital image noise filtering algorithm about median filtering and Wiener filter, the thesis introduces adaptive median filtering algorithm and improved Wiener filter algorithm to reduce image noise of high resolution x-ray gastrointestinal digital image and drawn relative resolution conclusion. As to requirements of filtering image noise of 2048×2048×12bit high resolution x-ray gastrointestinal digital image system, the thesis use wavelet transform for many research method of high resolution digital image denoising for reference, the thesis put forward a coefficient based on the classification of wavelet domain mix-model image denoising algorithm for high resolution x-ray gastrointestinal digital image denoising. This algorithm utilize wavelet zero tree structure to represent the correlation between scales, then classified wavelet modulus and adopt mixed-model approximation of part statistical characteristic of each type, realizing image noise reduction processing in the framework of Bayes. According to research of testing the simulation and the actual of high resolution x-ray gastrointestinal digital image, it shows that the algorithm could effectively reduce noise while preserve detailed info of image, overcame the deficiencies of median filtering and Wiener filtering traditional digital image noise filtering algorithm easy to create image detail fuzzy, it provide reliable guarantee to doctor's accurate diagnosis of gastrointestinal symptoms.