Research On Noise Reduction And Enhancement For Medical X-ray Images

Due to the low cost, high speed and high reliability, X-ray transillumination imaging has been an important method of routine physical examination. However, because of the complexity of body tissue, the effects of X-ray scattering and electrical noise etc. the X-ray medical image which is obtained directly by the sensor has low signal-to-noise ratio and low contrast, it can’t be a diagnosis reference for the doctors yet since we do the image processing.The conventional noise reduction algorithm of X-ray medical image does not consider the differences in the noise type and the image difference between the different body parts. Along with the noise reduction, the useful information is also blurred. It is also a lack of considerations for the structure and density properties of complexity and the different parts of the body of the structure of the human body in image enhancement. The treatment effect on special object or image is far from ideal. Noise reduction and image enhancement algorithm remains to be further improved. In this paper, combined with the domestic X-ray machine in the development of image processing research, we do the research and the software development for the noise reduction, image enhancement, gray-scale transformation.Firstly, combined with the development of X-ray machines through the experiment and research on the X-ray machine, we analyzed the cause reasons of the X-ray medical original image noise and type. Analyzed the distribution of noise, analyzed and judge the noise type of X-ray medical images according to the distribution of noise, so we can more focused on noise reduction processing.Then we designed three kinds of improved noise reduction algorithm for the different types of X-ray medical images. The improved median filter algorithm for impulse noise distinguishes the signal and noise firstly, then just processing noise point without damaging the signal point. the improved median filter algorithm for the Gaussian noise we adopt the method of subsection statistics, Only take the middle segment involved in processing to reduce the image blur. For the quantum noise we using the wavelet threshold noise reduction algorithm, Combined with the hard threshold and soft threshold method for processing, And obtained more effective noise reduction affection.We combined with the X-ray medical examination for the different parts of the body and the requirements. In this paper, we combined with the different inspection area to design the adaptive parameter setting method for the X-ray medical image enhancement processing. By means of contrast enhancement, edge and detail enhancement, low contrast and fuzzy details of the original image is improved. When display window gray adjustment, for the different parts of the human body the range value of the window wide is different. This paper put forward in X medical image by using the adaptive enhancement method, considering the X-ray medical images of different density in different parts of human body in image enhancement processing, adaptive window adjusting processing parameters, obtained better effect.Finally, experiments and result analysis have been taken for the designed image noise reduction and image enhancement methods on the X-ray medical image. Experimental results show that the improved algorithm design both from the subjective and objective indicators have achieved better effects. The algorithm of this study has been carried out in software development and application of newly developed X-ray machine-made device, and achieved good results on X-ray medical image processing.