快速有效的全局和局部能量驱动的活动轮廓模型

Fast and Efficient Active Contours Driven by Local and Global Intensity Fitting Energy

由全局和局部拟合能量驱动的活动轮廓模型(LGIF模型)对活动轮廓的初始化和噪声不敏感,且能够分割灰度不均匀图像;但是该模型的演化方程在每次迭代中需要进行多次高斯卷积,使得分割速度非常慢;基于这一缺点提出了一个新的模型;实验表明:该模型不仅能够分割灰度不均匀图像,而且分割效率优于LGIF模型。

Active contours driven by local and global intensity fitting energy （LGIF model） is much less sensitive to the initialization of the contours and noise, and it can address the segmentation of images with intensity inhomogeneity. However, the evolution equation of this model needs many Gaussian convolutions per iteration, which make the segmentation speed extremely slow. We propose an improved model based on the aforementioned fault. The experiments show that the model is not only able to deal with intensity inhomogeneity, but also the segmentation efficiency is more efficient than LGIF model.