图象相位特征提取及其在纹理分割中的应用

On Improving Texture Segmentation

强调了长期为人们所忽略的相位信息在纹理分割中的重要性，在介绍相位缠卷概念的基础上给出了盲目展开、胞腔式自动机、相位解调等几种获取局部非缠卷相位的方案，并对它们各自的优缺点进行了简单的阐述，最后给出了将展开相位作为特征信息应用于纹理分割的实例。

To our best knowledge, existing texture segmentation methods Cart not succeed in distinguishing between two neighboring regions with the same multitude but different phasesowing to the neglect of full exploitation of phase information. We now abc to achieve suchdistinguishing.Eqs. (14) and (15) can be used to unwrap phases, which, in turn, distinguish betweenneighboring regions which differ only in phase. Eqs. (14) and (15) are well known to thosefamillar with blind unwrapping method [1], but to obtain better results, we make use of ourpast experience in processing with Eqs. (14) and (15).In our one experiment, we use orientational fliters to extract multitude and wrappedphase features. Then blind unwrapping method is used to obtain unwrapped phase. Fig' 1(a)is the original texture image, Fig. 1 (b) is the segmentation result using only multitude features, and Fig. 1(c) is the result obtained by sdding phase information. We can see that ourmethod (Fig. 1c) can easily distinguish between two neighboring regions which differ only inphase.