基于分形与灰度特征的无监督纹理分割技术

Unsupervised texture segmentation by integration of fractal and grey-level features

提出了一种新的基于方向分形特征和灰度特征的纹理图像分割方法。该方法首先用一个局部窗从功率谱图像中提取不同方向上的分形维和分形截距,将它们各自的均值和方差与灰度均值、灰度方差结合起来构成一个多维特征向量,然后利用模糊C均值聚类算法进行聚类实现纹理图像的分割。实验结果表明该方法对织物纹理图像和医学图像都有着良好的分割效果,鲁棒性强。

A new texture segmentation method based on directional fractal values and gray-level features is proposed in this paper.Firstly,the directional fractal dimensions and corresponding intercepts are extracted from a power spectrum image with a given local window,and their respective statistics（mean and variance） are combined with gray-level mean and variance to form a multidimensional feature vector.Then we apply the fuzzy c-means algorithm as the classifier based on formed feature vectors to get the texture segmentation result.The experiment shows that the proposed method is significantly effective and robust both on textile and medical images.