无须反复初始化的活动围道纹理分割方法

Texture image segmentation using the without re-initialization geodesic active contour model

针对纹理图像分割问题,提出了一种无须反复初始化的快速活动围道纹理分割模型.与现有的基于几何活动围道的纹理分割技术相比,该分割模型具有三个优点:一是使用图像的灰度信息和纹理信息来驱动活动围道进行图像分割,因而,不仅适用于分割纹理图像,而且适用于分割非纹理图像;二是使用局部二进制模式来提取纹理特征,计算复杂度小;三是模型求解时,增加了约束项不需要对符号距离函数进行反复初始化,因此可采用大的时间步长,迭代步数明显减少,从而提高了活动围道的收敛速度.对自然界真实图像和合成纹理图像的分割试验结果说明,无须反复初始化的活动围道纹理分割方法精度高、速度快.

Segmenting an image into differently textured regions is a difficult problem. A new method for texture image segmentation is proposed, which has three advantages over the other active contours. Firstly, by combining the gray levels of pixels and texture information of an image, this method can be used for segmentation of a texture image or a none-texture image. Secondly, the method has low computation complexity, because the LBP （local binary pattern） is employed to extract texture features. Finally, the without re-initialization algorithm proposed in this paper can avoid the additional computation problem due to the re- initialization of the signal distance function. The segmentation tests for synthetic and natural texture images show that the proposed segmentation method is efficient, accurate, fast and robust.