首页 | 本学科首页   官方微博 | 高级检索  
     检索      

改进Tamura纹理特征的图像检索方法
引用本文:郝玉保,王仁礼,马军,苏斌,郑建华.改进Tamura纹理特征的图像检索方法[J].测绘科学,2010,35(4):136-138,176.
作者姓名:郝玉保  王仁礼  马军  苏斌  郑建华
作者单位:信息工程大学测绘学院,郑州,450052;75719部队,武汉,430074;河海大学水资源环境学院,南京,210098;75719部队,武汉,430074
摘    要:针对Tamura方向度纹理特征计算时直方图峰值难以确定的问题,本文提出了采用改造后的旋转不变边缘方向直方图代替方向度纹理特征;针对粗糙度纹理特征不能充分反映图像上纹理基元尺寸的大小分布的问题,提出了采用粗糙度直方图代替粗糙度的特征表达方式。实验表明,改进后的Tamura纹理特征易于获取,包含图像更多的信息,显著地提高了检索性能。

关 键 词:Tamura  纹理特征  粗糙度直方图  边缘方向直方图  图像检索

Image retrieval based on improved Tamura texture features
HAO Yu-bao,WANG Ren-li,MA Jun,SU Bin,ZHENG Jian-hua.Image retrieval based on improved Tamura texture features[J].Science of Surveying and Mapping,2010,35(4):136-138,176.
Authors:HAO Yu-bao  WANG Ren-li  MA Jun  SU Bin  ZHENG Jian-hua
Abstract:For the peak of histogram is difficult to be ascertained when the texture feature of Tamura directionality is computed,the authors replaced the directionality texture feature with the revised edge directional histogram invariant to rotate.Meanwhile,for the texture feature of Tamura coarseness can't respond to the changes in size of texture primitive adequately,the authors replaced the coarseness with the coarseness histogram.The experiments performed on the Brodatz texture database and remote sensing images indicated that the improved Tamura texture feature could be attained easily,contain more valuable information and increase the precision of retrieval remarkably.
Keywords:Tamura
本文献已被 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号