小波变换在纹理提取技术中的研究现状  

Wavelet Transform in Texture Extraction Technology Research

作　　者：李一波[1] 朴慧[1] 

机构地区：[1]沈阳航空工业学院自动化学院,辽宁沈阳110136

出　　处：《小型微型计算机系统》2010年第6期1213-1218,共6页Journal of Chinese Computer Systems

基　　金：沈阳市人才专项无人飞机遥现操控与模拟系统研制项目(06syrc03)资助

摘　　要：遥感技术的发展带动了图像处理技术的不断进步和发展.遥感图像因其特有的图像特点,使得地物在空间上的排列更加的复杂多变.目前,小波技术因其本身"多分辨率"的特性,能在时频两域都具有表征信号局部特征的能力和多分辨率分析的特点,因而被广泛地应用到图像处理的各个领域中去.本文将近几年小波在图像处理领域中应用比较多的三种处理方式进行了总结和比较,最终认为基于复小波的处理方式更加的精确,合理,在此基础上提出要更多的应用复小波方式处理图像,以及不再采用单一的处理方式,而是融合各种小波方式,结合各种方式方法的优点,使图像处理的更加清晰,准确.The development of remote sensing technology drives image processing research undergoing continuous development and progress.Because of its peculiar image characteristic,remote sensing diversifies the arrangement of the objects and the land in the space.At present,due to its own property of ＂multi-resolution＂ as well as its capacity of behaving the signal＇s local features and of multi-resolution analyzing,the wavelet technology has been widely used in every domain of image processing.In this paper,three wavelet processing modes which have been widely used in the domain of image processing have been summarized and compared.It is found that the processing approach which based on complex wavelet transform is more precise and reasonable.So the more complex wavelet technology has been strongly recommended in image processing.To integrate all sorts of wavelet technologies and their advantages rather than adopt single processing approach would make the image processing more vivid and accurate.

关 键 词：小波变换 纹理特征 多分辨率 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]