摘要
为使图像处理更快速稳定,文中通过研究偏微分方程的元胞自动机模型,提出了一种基于格子波尔兹曼模型(LBM)对流扩散方程的图像分割算法:用对流扩散方程替换水平集方程,利用LBM求解对流扩散方程,实现基于LBM的水平集图像分割。实验分析表明,该方法能精确求解对流扩散方程,算法模型能够取得闭合的分割曲线,同样能够很好的处理图像拓扑结构的变化。并且通过与水平集方法和窄带水平集方法进行计算速度实验对比,可证明该算法大大减少了分割的计算量。
In order to make the image processing faster and more stability, the cellular automaton model founded on partial differential equations is studied. Based on lattice bohzmann model(LBM) , a method of image segmentation by convection-diffusion equations is proposed in this paper. Firstly, the level set equation is replaced with convection-diffusion equation. Then, it used of LBM solving convection-diffusion equation to achieve the level set of image segmentation. The experiment result shows that our algorithm can accurately solve the convection-diffusion equation. It can obtain closed curves and deal with topology changes well. Contrasting with the level set method and narrowband level set method for computing speed, the result of the experiment is confirmed that the algorithm greatly reduces the segmentation of the computation.
出处
《信息化研究》
2010年第7期29-33,共5页
INFORMATIZATION RESEARCH
关键词
图像分割
格子波尔兹曼模型
对流扩散方程
水平集方法
image segmentation
lattice Bohzmann model
convection-diffusion equations
level set method