Abstract: This paper puts forward the base frame of lung segmentation first.Using optimal thresholding and mathematical morphologic methods，the rough image of lung is acquired.Then，we present a fast adaptive segmentation refinement algorithm，aiming at the unsuccessful left and right segmentation of thredsholding.Experiments on clinical 3D pulmonary images show the algorithm’s result is very close to the expert manual outlines，and it is very effective for the separation of left and right lung with a successful segmentation ratio 94.8%.

Key words: image segmentation, CAD system, pulmonary CT, automated thresholding, watershed segmentation

摘要： 首先给出了一个肺实质分割的基本框架，结合最佳阈值法、数学形态学方法，对图像进行了粗分割；然后，针对左右肺未完全分离的情况，提出了快速自适应的优化细分割方法。在实际临床胸部CT图像上的实验表明，该方法的分割结果和专家的人工分割结果很接近，特别是对于左右肺的分离有很好的实验效果，成功分割率达到94.8%。

关键词: 图像分割, CAD系统, 肺部CT, 自动阈值, 分水岭分割