节点文献
基于颜色和纹理特征的图像检索技术研究
Image retrieval based on color and texture
【摘要】 本文提出一种基于颜色和纹理特征的图像检索方法,利用颜色直方图表示图像的颜色特征,计算图像之间的颜色距离;在提取图像的纹理特征时,首先对原图像进行压缩和灰度化,然后利用图像的灰度共生矩阵,统计出反映纹理一致性、纹理对比性、纹理的嫡和纹理相关性的数据。最后利用这两个特征进行图像检索。实验表明,本文的算法具有较好的检索效果。
【Abstract】 A new method using color feature and texture feature is proposed.Firstly the color histogram is extracted,then the distance is computed.In order to extracte texture feature,the image is compressed and convert it to gray,then the four numercial features identifying the content of image are got from gray-level cooccurrence matrix of the image.At last,according color feature and texture the similarity is analyzed.It is demonstrated that the methodology is feasible.
【关键词】 图像检索;
特征提取;
颜色特征;
纹理特征;
相似度量;
【Key words】 image retrieval; feature extraction; color feature; texture feature; similarity matching;
【Key words】 image retrieval; feature extraction; color feature; texture feature; similarity matching;
- 【文献出处】 中国科技信息 ,China Science and Technology Information , 编辑部邮箱 ,2008年08期
- 【分类号】TP391.3
- 【被引频次】5
- 【下载频次】360