Based On The Optimization Of Watershed And 3 D Reconstruction Of Medical Image Segmentation

3D reconstruction technology of medical image shows three-dimensional graphics data to thediagnosis of personnel, in order to facilitate the data on patients of multi-directional and multi-levelobservation, and qualitative, quantitative analysis is completed by the doctor about diseased tissues.Research contents of3D reconstruction of medical image include image preprocessing,segmentation and extraction of tissue or organ, reconstruction algorithm research.The imagesegmentation is a very important step, and effect of the segmentation results have a direct impact onthe follow-up processing efficiency and quality. There are many kinds of image segmentationalgorithm, in which watershed algorithm calculation speed, accurate positioning and the contourline is closed, so it has been widely applied. But the watershed algorithm also has a serious problemthat is easily over segmentation.The watershed segmentation algorithm is very easy to appear the phenomenon of oversegmentation, so the traditional watershed algorithm is improved. This paper puts forward a newwatershed method for medical image segmentation based on threshold marking. The originalcluster image processing, and maximum entropy threshold method is used to extract the internalmarker, and watershed transform are used on internal markers to get external marker. Modifygradient map with internal and external marker maps and use watershed transform on the modifiedgradient map.Realization of discovery, the method not only overcomes the artificial extractionmark difficulties, but also effectively solve the over segmentation of watershed algorithm. Finally,the segmentation results is subsequently processed, to achieve the data processed3D surfacereconstruction though the MC algorithm, and in the MITK platform to achieve a3D image datameasuring and cutting and other interactive operation.