Algorithm For Cell Image Segmentation Based On Mathematical Morphology

With the development of science and technology, especially the information technology, image processing technology has become one of the fundamental tools on scientific research. Image segmentation, a key technology of image processing, develops rapidly in the recent years and deeps into the many aspects including industry, military affairs, medicine, spaceflight, etc. And its applications could be found easily in above fields.Medicine Imaging Processing is an idiographic application about the Digital Image Processing technology, and it has been already regarded more diffusely in all over the world. The segmentation of cell image, one of the important embranchment of Medicine Imaging Processing, comes in for extensive attentions by so many computer image processing researchers. The segmentation and recognition algorithm of cell images, progressed with the development of computer technology, has turned into the one-up research task in the present image domain.As far as I'm concerned, the study about the cell image processing includes the following parts:1. Image Pretreatment. The pretreatment technologies about the image processing are composed of image enhancement, median filter, edge detection and so on. How to apply these methods reasonably and how to combine these methods efficaciously is the sticking point of obtaining the good pretreatment results. By repetitious experiments, results analysis and contrast, an effective pretreatment means for images of blood corpuscle is get, and this means establishes favorable base for the image segmentation.2. Morphological Mathematics Theory. It's commonly used at the Medicine Imaging Processing. So the learning emphasis of this part is to apply the theory to the cell images, observe what results will be brought out, analyze the used occasion of the methods in this theory.3. Cell Image Segmentation. Image segmentation is involved in expansive ken, so it's one of the difficulties in the learning process. The experiments adopt the following methods to segment the cell image and compare the segmentation results: threshold segmentation, region growth technology, also as some other way developed in the recent years, like the segmentation technology based on the training of gray swatches, the object delineation algorithm base on the gray density.4. Watershed Segmentation Algorithm. The core of this part is to improve the traditional watershed algorithm according to the characteristic of blood corpuscle image. Through the pretreatment in prophase and the catchment basin chosen, the over-segmentation phenomena existing in traditional watershed algorithm can be restrained. Experimental results show that the proposed algorithm can segment images of blood corpuscle quickly and accurately, and exactly locate edges. Furthermore, the influence of noisy can be averted by this way.5. Experimental Conclusion. At this part, compare and analyze the experimental results, explain the key and the difficulty of this subject at the next work step, also, take an expectation on the development of cell image processing.