Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (30): 188-193.DOI: 10.3778/j.issn.1002-8331.2009.30.057

• 图形、图像、模式识别 • Previous Articles     Next Articles

Finite texture mixture model and corresponding texture segmentation scheme

XIE Xing,XIE Yu-bo,QIN Qian-qing   

  1. State Key Laboratory of Information Engineering in Surveying,Mapping and Remote Sensing,Wuhan University,Wuhan 430079,China
  • Received:2008-11-18 Revised:2009-01-14 Online:2009-10-21 Published:2009-10-21
  • Contact: XIE Xing

有限混合纹理模式及其纹理分割框架

谢 兴,谢玉波,秦前清   

  1. 武汉大学 测绘遥感信息工程国家重点实验室,武汉 430079
  • 通讯作者: 谢 兴

Abstract: With the illumination of the texture analysis methods based on model and structure,a new feature extraction method,Finite Texture Mixture Model(FTMM) is proposed in this paper.Because of the calculation of the texture feature is found on the clustering methods,the complex process of parameter evaluation of model-based methods and the deficient ability to describe texture of structure-based methods can be avoided.Based on FTMM,the corresponding texture segmentation scheme is given,and the effectiveness of the method is proven by quantitative and qualitative experiments.

Key words: Finite Texture Mixture Model(FTMM), local variable model, texture segmentation

摘要: 受到基于模型和基于结构的纹理分析方法的启发,提出一种新的特征提取方法—有限混合纹理模式。纹理特征采用聚类的方法进行计算,避免了基于模型方法复杂的参数求取过程,同时也突破了基于结构的方法纹理表达能力不足的问题。在该特征的基础上,给出相应的纹理分割框架,并通过定量和定性实验验证了所提算法的有效性。

关键词: 有限混合纹理模式, 局部变化模式, 纹理分割

CLC Number: