CT图像椎骨分割的局部模糊主动轮廓方法

摘要

针对医学CT图像灰度值分布不均匀的问题，由于模糊能量主动轮廓（FEAC）模型用全局灰度信息对图像进行分割，导致过分割现象。对该方法作出改进，即一种局部化FEAC（LFEAC）模型，利用局部的加权平均值代替全局的均值。将该方法用于医学CT图像椎骨分割，实验表明分割效果比FEAC模型和CV模型好。%The gray values of the medical CT image are uneven distributed. The curve evolution is based on the global gray value information in the Fuzzy Energy-based Active Contours(FEAC)model. So this model will result in over-segmentation phenom-enon. This paper presents an improved model for FEAC. That is localized for FEAC(LFEAC). In LFEAC model, instead of global gray value information it uses the local gray value information. This paper segments the medical CT vertebral with the method of LFEAC, and gets a satisfied result. The result is better than the result of FEAC model and CV model.