Segmentation and Visualization using Chinese Visible Human
Abstract:

Background removal is a critical step in Visible Human Data (VHD) processing, which is the basic of all other researches. A new segmentation algorithm based on hybrid method for VHD background removal has been proposed. The method combines a feature based segmentation method, using the color features of the background and the interested parts, with a contour based one, the Gradient Vector Flow (GVF) Snake model. Our testing results show that the new algorithm is more robust and highly accurate compared with previous methods.

Since visible human visualization using cryosection images is still a challenge for its own difficulties such as color inhomogeneity between adjacent images, most visible human visualizations use pseudo color. We propose a method to reconstruct and visualize 3D visible human head with an approximate and reasonable real surface color. Some key algorithms are presented, including color homogenization, model creation and color generation methods. These algorithms are implemented by using GPU, and the 3D head model can be rendered in real time. The experiment on head data of Chinese Visible Human shows that our method is successful for the 3D colored reconstruction and visualization.


References:

  1. Chen Ding ,Yankui Sun,Tiaolin Tian,and Zesheng Tang,A Hybrid Method for Automatic and Highly Precise VHD Background Removal. Lecture Notes in Computer Science, "Medical Imaging and Informatics". Volume 4987,294-303,2008

  2. Fan Bao,Yankui Sun,Tiaolin Tian,and Zesheng Tang,3D Head Reconstruction and Color Visualization of Chinese Visible Human. Lecture Notes in Computer Science, "Medical Imaging and Informatics". Volume 4987,262-269,2008
[ ]