The technology of image mining is an important part of technology of data mining, it has get attention, going with the continuous development of data mining. At present, the image database of world, many of the images were produced one day. but for the vast amounts of data, we do not know how to deal with it. The technology of image mining should be developed quickly. This article describes the general method of image mining and its application to the classification of brain CT images, get much more accuracy. Now image mining was application in medicine at an early stage, and applied to biomedical. medical image of breast cancer of analysis of disease, analysis of liver medical image and analysis of sequence of DNA gene in the past, and the analysis of cardiogram map, the research in these areas have a precedent, and also to apply to the practice.This article describes much of the medical data mining-related knowledge. The data of medical image come from brain CT images, firstly analyzing various factors to the impact of quality of CT image and corresponding image preprocessing based on these factors. Aim reducing the image noise, and further enhance the image. In this paper, the second counterchange of the Laplace transform of the image are enhanced.After dealing with the image preprocessing, in order to describe the character information of image, making us understand the information of image. We should carry out the feature of image extraction, to get features from the image information. Describe the information of image base on the image color, texture, shape and other characteristics. Mainly base on color feature extraction method using a mean of Histogram, and according to get some mathematical statistics to account the color information characteristics of image. Acquire texture of feature extraction base on gray level co-occurrence Matrix analysis. In this paper, analyze the shape feature of the image the by method of Fourier descriptors.The last part of paper is mainly narrate the classification of data mining base on the extracted image features, the paper used CART decision tree algorithm for classify image of medical, and the algorithm base on the betterment of CART algorithm. And doing experiment base on the database of medical image,the brain CT images come from the third layer to the top of the brain.analyzing experiment and showing the good degree of correct rate of classification. |