Research On Watershed Segmentation Algorithm Of Color Image Based On Morphological Gradient

Image segmentation is a key step from image processing to image analysis.So far,the watershed algorithm not only overcomes some shortcomings and drawbacks of the traditional image segmentation algorithm,but also has a short computation time.Therefore,the algorithm is one of the algorithms of segmentation algorithm.The watershed segmentation method is an effective image segmentation method,but it will produce the phenomenon of over-segmentation.In view of this problem,the paper divides the watershed segmentation algorithm with color space separately,and obtains two color image watershed segmentation algorithms.It is applied to the image analysis of the anterior segment of medicine to further improve the practical use value of color image segmentation.First of all,the paper introduces the background and importance of image segmentation,and summarizes the research status and research significance of aggregate watershed algorithm and medical image segmentation.And introduces the definition,nature and geometric example of mathematical morphology.The principle,definition and algorithm of watershed segmentation are expounded in detail,which paves the way for the algorithm and application of later chapters.Secondly,in the RGB color space,the watershed algorithm of new morphological gradient is developed.In view of the problem of over-segmentation and edge detail loss caused by watershed algorithm,combined with the advantages of morphological gradient and information entropy,using the advantages of regional merging,Not only solve the loss of color information caused by segmentation,image over-segmentation problem,and the algorithm is simple and very good to speed up the split speed.Thirdly,in the HSI color space,a watershed segmentation algorithm based on the flexible mathematical morphology is proposed.The advantages of the anti-noise of the flexible morphological operator and the loss of the color information when the color image is transformed into the gray scale image Problem,you can not turn gray image directly on the color image segmentation to get color gradient.Finally,the algorithm is applied to the image of the anterior segment of the medicine,and the corresponding image of the anterior segment of the medical image is selected.The experimental results show that the image segmentation algorithm can help the medical staff to further diagnose and treat.