基于改进型C-V模型的红外图像自动分割

将C-V模型引入到红外图像分割中,同时提出了一种改进的模型。该模型通过增加偏离能量项,避免了标准C-V模型的重新初始问题。模型允许选取较大的时间步长,可以用简单的有限差分格式实现。实验证明,改进的模型在分割效果不变的情况下可以有效加速演化过程,提高分割效率。

Automatic Segmentation of Infrared Images Based on an Improved C-V Model

The C-V model is introduced into the infrared image segmentation,and an improved C-V model is proposed.This improved model avoids the problem of re-initialization in the normal C-V model by adding a deviatory term into the C-V model.It can be implemented by simple finite difference scheme with a significantly larger time step.The experimental results show that the improved C-V model can speed up the evolution and make the segmentation more efficient while there is no degradation in the segmentation result.