摘要
基于内容的图像检索是数字图书馆和多媒体数据库的重要研究内容。文中主要针对图像处理,特别是运用图像的纹理特性来查找或检索图像数据库。运用Gabor 变换进行纹理分析是比较成功的方法。小波变换尽管在这方面也取得了较好的效果,但在许多情况下,不如用Gabor滤波器效果好。文中就是在Gabor 滤波器的基础上生成Gabor 小波,对图像进行多分辨率分析,提取纹理特征,提供了一种基于图像纹理的图像检索方法,并对此给出了一些实验结果。
Image retrieval based on content is an important study aspect in digital library and multimedia database.In this paper,image processing especially at retrieval image database using picture′s texture characters is discussed.Using Gabor transformation to analyze picture is a successful method. Although wavelets transformation gets satisfactory results in many cases, Gabor filter is more effective than it.In this paper, the Gabor wavelets based on Gabor filters are generated to process picture with multi\|resolution analysis,and extract the image features with Gabor wavelets, and then give a method of image retrieval based on texture features. Some results of experiment are given in the end.$$$$
出处
《红外与激光工程》
EI
CSCD
1999年第6期25-28,57,共5页
Infrared and Laser Engineering
关键词
纹理特征
图像检索
GABOR函数
图像处理
Texture feature\ \ Image retrieval\ \ Gabor function\ \ Gabor wavelets