基于极坐标Log Gabor小波的纹理分析方法  

A METHOD FOR TEXTURE ANALYSIS BASED ON POLAR COORDINATE LOG GABOR WAVELETS

作　　者：刘琼[1] 周慧灿[1] 王耀南[2] 

机构地区：[1]湖南文理学院计算机科学系,湖南常德415000 [2]湖南大学电气与信息工程学院,湖南长沙410082

出　　处：《计算机应用与软件》2008年第8期234-236,共3页Computer Applications and Software

摘　　要：提出一种基于极坐标Log Gabor小波的纹理分析方法,该方法突破了Gabor小波的带宽限制,仅采用较少的极坐标Log Gabor滤波器,即实现了准确地纹理分割。实验表明,在频谱覆盖相同的情况下,极坐标Log Gabor小波具有较Gabor小波更好的性能和更高的效率;此外,实验中针对多通道滤波方法对最小距离判别方法的改进,也取得了较好的效果,这为基于小波分解的分类问题提供了一个新的思路。A method based on polar coordinate Log Gabor wavelet for texture analysis was brought forward. The method broke through the bandwidth limitation of Gabor wavelet, and realized texture segmentation with accuracy by only adopting a few polar coordinates Log Gabor filters. Experiments indicated that, covering the same frequency spectrum, polar coordinates Log Gabor wavelet has better performance and more efficiency than Gabor wavelet. On the other hand, an improvement to minimum distance classifier aiming at multi-channel filtering was made in experiments, and got a better result and also provided a new approach for classifying problems which are based on wavelet decompoing

关 键 词：GABOR小波 极坐标Log GABOR小波 纹理分析 最小距离判别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP316[自动化与计算机技术—计算机科学与技术]