Image segmentation by graph cut for radiation images of small animal blood vessels.

摘要: Synchrotron radiation (SR) based X-ray imaging is an attractive method for analyzing biomedical structure. However, despite its many advantages, there are few gold standards in image processing methods, especially segmentation. Image segmentation essential step medical analysis and three-dimensional reconstruction. Although algorithms segmentation, a decisive does not exist SR imagery, because of lack data. This study focused on finding suitable algorithm high-resolution imaging. In this study, we used following four to segment blood vessel mouse; interactive graph cuts algorithm, which segments using fast min-cut/max-flow solve global solution, binary partition tree uses creating nodes by splitting merging image, seeded region growing performs connecting similar pixel value, simple object extraction, generates color signature model image.