在最优阈值分割的基础上,用自动区域生长去除气管/支气管区域,对边界跟踪法进行改进以快速

去除背景干扰和获得肺部边界,最后进行肺部边界修补得到完整的肺部图像。算法采用迭代法寻找最

优阈值解决了阈值选取的敏感性问题,提出了基于前层图像中气管/支气管位置的气管/支气管提取方

法,避免了种子点的人工选取,基于前次搜索方向改进了八邻域搜索方法来提高边界跟踪的速度。

-In the optimal threshold segmentation based on region growing with automatic removal of the trachea/bronchus region, to improve the boundary tracing method to quickly remove the background interference and access to the lung border, and finally to get the complete repair of lung lung border image. Algorithm uses the iterative method to find the optimal threshold to solve the sensitivity threshold selection problem, a layer of the image based on the previous tracheal/bronchial location trachea/bronchus extraction method, to avoid the manual selection of seed points, based on previous search the direction of improving the eight neighborhood search methods to improve the speed of boundary tracking.