能量自适应的快速水平集图像分割模型

A fast-level set image segmentation method with adaptive energy

针对灰度不均匀图像和噪声图像的分割难问题,本文提出一种高斯窗口和权重自适应的快速水平集图像分割模型。首先,融入新边缘指示函数生成全局能量项,利用新边缘指示函数构造自适应高斯核函数,并由此构建局部能量项;其次,利用演化曲线内外灰度拟合均值差和局部可变区域灰度均值差构造自适应权控函数,使模型能自适应地调节全局力和局部力的比重;最后,采用有限差分法进行数值求解。实验结果表明,本文模型能有效分割灰度不均匀图像和噪声图像,提高了分割效率和对初始轮廓的鲁棒性。

A fast-level set image segmentation model-method with adaptive energy is proposed for the segmenting of gray scale inhomogeneous images and noisy images.Firstly,the global energy term is generated by incorporating the new edge indicator function in the global fitting term.The adaptive Gaussian kernel function is constructed using the new edge indicator function.constructing the local energy term.The adaptive weight control function is constructed using the difference between the mean value of the gray scale fitting inside and outside the evolution curve and the mean value of the gray scale difference in the local variable region.The model can thus adjust the weight of the global force and the local force adaptively.Finally,the finite difference method is used for the numerical solution.The results show that the model in this paper can effectively segment gray scale inhomogeneous images and noisy images.The model can also improve the segmentation efficiency and robustness to the initial contours.