Lung CT Image Segmentation Based on Combined Multi-Scales Watershed Method and Region Growing Method
Juanjuan Zhao, Jianfeng Yang, Yan Qiang, Quan Wang · International Review on Computers and Software (IRECOS) · 2013
In the field of computer assisted diagnostic, Image segmentation is one of the most important preprocessing steps. In this paper, the watershed algorithm segmented image by choosing different merging parameters, original image will be divided into different scales of image, High-scale image and low-scale image. High-scale image is part of the pulmonary parenchyma without the main vessel in it, low-scale image used to aid region growing tore move the branching vessels from the pulmonary parenchyma. Selected seed point located inside the branch vessels and determined the region growing threshold, with the aid of low-scale image region growing can segment the branch vessel regions .Removed the regions in the high-scale image to get the final segmentation results. In this paper, a large number of CT lung data segmentation experiment. Proven, good segmentation results are achieved.