In this paper, the division of pulmonary capillary extraction region, according to the lung itself, the gray image information, and capillary own characteristics, such as the gradual change in gray, uneven thickness, the main advantage of region growing method, Laplace second order differential operators and some other image processing algorithms, the more complicated parts of the extraction capillary, this method overcomes the shortcomings of a single algorithm to effectively separate out the shape and direction of major blood vessels, has proved a certain validity and application.