基于NBS颜色距离及半监督聚类的织物图像分割

Textile Image Segmentation Based on Semi-supervised Cluster and NBS Color Distance

为提高纺织CAD技术,对织物图像的分割进行了研究。依据半监督聚类理论,提出了一种基于HLC颜色空间的以NBS颜色距离为优化准则的半监督聚类的织物图像分割算法。算法利用有限的人工信息,即在织物图像上点击有限的几个点以标识相应区域之间的关系,从而得到满足用户给定限制的织物图像分割结果。算法首先对织物图像进行量化转换处理,而后在HLC色彩空间中集成先验的分割信息进行色彩聚类。实验结果表明,该算法在织物图像分割中是一种可行的方法。

To improve the technology of the textile CAD,segmentation ways of textile image are analyzed.Based on the semi-supervised clustering,proposed a new way of textile image segmentation by the NBS color distance in HLC color space.The algorithm makes use of limited human assistance,i.e.by indicating the relationship of some different regions in the textile image by mouse,to get the final accurate results of segmentation.The algorithm quantizes the textile image.And it then clusters colors in the HLC color space with prior segmentation information.