| As an important application field of image processing technology, medical image processing has been attached importance widely by pursuers in the field of digital image processing, along with the development of digital image processing, researches of methods on medical image processing and recognition also become a front problem of present medical image domain. The development of image processing and analysis and its application on medical images has provided more abundant and exacter medical information. In medical diagnose, because of high occurrence rate and high death rate, cancer diagnose has become a research hotspot of it and the research focuses on the problem of recognition of cancer cell, domestic and foreign medicinal image researchers have raised many effective image processing and recognition methods in recent years, The cancer cell auto-recognition systems based on these theories and methods are very important in computer assistant diagnosis, especially in the instances of lacking experts.In the field of medical image processing, it's an important application of image processing linked to the characteristics of medical images to carry out image processing and analysis through computer. After studying the technology direction and development of this field inside and outside our country, the author sum up primary research content and methods and introduce the research stress and kernel techniques, for instance, information fusion, mathematic morphology, image segmentation, etc. Based on these theories and techniques, this paper made a series of process on images combined with the characteristics of medical images, adopted appropriate processing and got better results by combining these techniques.In the field of medical cell image processing, segmentation and recognition of cell in the image is an important and difficult aspect, resolve of cell region segmentation has undeniable influence on disease diagnosis,quantitative analysis of cell information and cell aberrance, especially in cancer diagnose. It's frequent that there are many cells in one medical cell image, we carry out feature extraction from each cell and then get all distinguish results of the whole image, however, so it's very important to get the region segmentation of each cell. This paper presents a new image region segmentation based on the search path of tree in data structure, that is, T-shape search segmentation, the whole process which contains preprocessing, feature extraction, classification and recognition is also introduced, in the end, that's a brief summarize of medical image processing and the research on cancer cell auto-recognition systems.According to the general process of medical cell images, we arrange contents of the paper as follows:In the introduction, the paper introduces the survey and development of medical digital imaging, feature analysis of cancer cells and cancer diagnosis, significance of the new problem is also explained.In chapter 2, that is, preprocessing phrase, firstly, the paper introduces briefly the acquisition of medical images, then common preprocessing methods in medical image processing, in the end, the paper advances image fusion arithmetic based on generalized images that aim at cell images.In chapter 3, aiming at the mathematic morphology theory used commonly in medical image processing and analysis, the paper introduces binary image processing which is correlative to it, the theory and medical cell image processing based on mathematic morphology.Chapter 4 is most important in the paper and also the hard core of cellimage processing------ image segmentation, this chapter advances a newcell region segmentation arithmetic based on data structure, that is, T-shape region segmentation arithmetic which is involved in some knowledge about data structure and joint region. This paper introduces the theory of data structure to cell image region segmentation and separation of cells is carried out, it enriches theory and technique of image segmentation.Chapter 5 introduces... |
PDF Full Text Request