Medicine HRCT Image Enhancement Based On Granular Computing

Granular Computing (GrC) theory, which mainly used in dealing with the large number of incomplete and uncertain problems among different grain levels, has become an hot issue in research area. The special tasks of GrC can be reflected in two aspects:Firstly, to reach a comprehensive cognition at a higher-level for various fields according to the commonness of the granular, without considering the distinctions of the low-level. Secondly, to conclude the theory of two respective independent areas between specific areas and particular areas under implicit structure which has been explicated. Based on multiple perspectives, GrC has certain independence by solving the complicated problems with systematic and structural approaches and, it also focuses on the comprehension of each perspective and researches of hierarchical levels.In order to achieve a satisfactory result and make it convenient for understanding and distinction, as one of the most basic image methods, Image Enhancement aims enhancing the useful information of the images by highlighting the detailed information in interested regions, as well as removing and reducing the unimportant or unnecessary information and areas. The main purpose of the enhancement is:Firstly, to improve the visual effects though the distinct image enhanced by a variety of techniques. Secondly, to convert the image into some forms which is convenient for analysis interpretation between human beings and machine. Nevertheless, Image Enhancement could just deal with the recognizing capacity for parts of the information in order to improve the application value of the image.As the theoretical basis among various theories of spatial processing technology, Image Enhancement applied a simple and effective technique, histogram, into the process of dealing with the images, while there exists a deficiency of lack of the detailed information. This thesis proposed a technique of Multi-histogram Equalization. It divides the original images, which is regarded as grains of knowledge, into several images according to GrC theory and use Histogram Equalization method to deal with the image grains which are proceeded to combine together. The purpose is to enhance the contrast and increase the detailed futures of the images, among which the most important is the satisfactory result that there is no lack of gray level.As refers to grayscale images, there always contain lots of information with the results displayed are no less near the same. Generally speaking, people’s ability to distinguish gray level can just reach dozens of levels and unable to extract such information in the image, while hundreds or even thousands of times than grayscale in distinguishing the luminance and chroma. Consequently, the introduction of Pseudo-color Enhancement technology improves both the visual effect and usage value of the images after dealing with the images with Multi-histogram Equalization technology.By adopting the methods discussed above into the HRCT images in chest diagnosis, this thesis conducted a large number of experiments by comparing with traditional enhancement methods. The results show that the medical HRCT image enhancement method based on GrC in this thesis has more advantages than the traditional. It is more suitable in dealing and analyzing the images in astrology and may bring much convenience for diagnosis and researches. Meanwhile, the display quality the medical images must be improved remarkably that will be convenient for doctors in their diagnosis and researches.