Skeleton extraction of color cerbrovascular image based on topological nodes

In this paper, a new continuous skeleton extraction method on color image was proposed. This method uses a new level set model in color space to extract the topology node of an object using a skeleton point as source-point which spreading a moderate wave-front, and then using these topology nodes as new source nodes, it spreads the second high-speed transmission wave front in order to identify skeleton points. All the parameters are obtained from the analysis, avoiding human intervention. Experiment and analysis can verify the validity and the robustness of this method, and it is insensitive to color gradient and boundary noise.