Abstract: The traditional methods for the segmentation and detection of pulmonary nodules could not segment pulmonary nodules accurately, and also could not separate the pulmonary nodules from blood vessels and chest wall. An improved random walker algorithm was proposed to solve the above problems. Firstly, according to the probability that calculated by Dirichlet boundary condition, the image will be divided into three parts: objective region, background region and uncertain region. Euclidean distance was used to calculate the difference between the nodes in the uncertain region and the seed, which can be used to label nodes in the uncertain region. Secondly, a parabola between two points algorithm (PBTP) for the second segmentation was proposed to achieve the final image segmentation. The experimental results demonstrated that the proposed algorithm segmented the pulmonary nodules more accurately than that of using the traditional methods, which greatly improved the accuracy of analysis and identification for computer-aided diagnosis (CAD) of pulmonary nodules.