基于提升算法的图像小波特征提取分析

Analysis on Image Wavelet Feature Extraction Based on Lifting Algorithm

讨论小波变换以及提升小波变换的基本原理和变换过程。利用2种变换分别实现了对同一图像库的特征提取,并利用支持向量机(SVM)进行分类。结果表明,与传统的小波变换相比,提升小波变换对于图像特征的提取同样是有效的,另外由于其独特的变换构造,使得计算量大大降低,计算时间显著减少,具有很大的实用价值。

This paper discusses the basic principles and the process of wavelet transform and lifting wavelet transform. The two transform methods are used to achieve the feature extraction for the same image library and to accomplish the classification with the Support Vector Machine. The results show that the lifting wavelet transform extraction is equally effective as traditional wavelet transform. Otherwise, because of its unique transformation structure, it can make the calculation greatly reduced, significantly reduce the computation time, and it has very practical value.