Research On Medical Image Enhancement Based On High Dimensional Information Geometry

Nowadays, medical imaging is one of the most commonly used and effective technology to monitoring the health of person, and large numbers of disease can be detected through the method of medical imaging(such as X scan, CT, MRI and so on). However, due to various of factors, medical image has the problem of blur and low quality, which seriously affects the effectiveness of doctors' diagnosis. As a result, medical image processing emerges at the historic moment accordingly and medical image enhancement becomes one of the currently widely focused researches.The medical image enhancement method based on high-dimensional information geometry proposed in this paper is an innovative method for fuzzy image enhancement. This method adopts the concept of high dimensional information geometry theory which maps the image into a point in the high-dimensional space. With the “blurring-clarity-blurring-clarity” thought, this method can get ideal enhanced images by comparing the corresponding locations of various images with simple geometry analysis. According to the bionic pattern recognition theory proposed by Academician Wang Shoujue, the amount of information between the original image and processed image are the same, and the two images can be regarded as two homologous things. Therefore, the change between the two images should be a continuous gradually changing process, i.e., the two images stand for two different points in the high-dimensional space and there should exist a continuous hypercurve between the two points, and each point in the hypercurve can be treated as an image correspondingly with the gray value of each pixel in the image(or RGB mode) used as their coordinates. Assuming that the size of the image is M* N, then the image is a point in the M* N dimensional space. And the relationship between the images we focused on is equal to the relationship between the points of high-dimensional space.The medical image enhancement method based on high-dimensional information geometry proposed in this paper can relatively easily get ideal enhanced images by simple geometry analysis. Compared with other image enhancement methods, this is a major breakthrough, which has a great advantage in the code concise extent. What's more, the method proposed in this paper has better enhancement effects, faster computing speed, and higher efficiency. Moreover, this paper proposes a novel medical image lesion detection method based on high-dimensional information geometry. This method establishes the mathematical model of the projection line segment enhancement algorithm and explores to combine the two dimensional image with the three dimensional space, which can effectively enhance the image contrast and visual effects.