小波降噪与独立成分分析的高光谱特征提取研究
首发时间:2011-07-27
摘要:针对高光谱数据的特点,提出了一种基于小波降噪和独立成分分析(Independent Component Analysis,ICA)的特征提取方法。该方法充分利用小波变换的优势,对高光谱数据进行降噪处理,旨在解决ICA对噪声过于敏感的问题。对降噪前后的原始高光谱数据通过独立成分分析进行特征提取,并采用支持向量机进行分类试验,用以验证降噪前后ICA特征提取的效果。试验结果表明,此方法可以有效去除原始影像中的噪声,并提高高光谱数据的特征提取和分类的精度。
For information in English, please click here
Wavelet-Based Independent Components Analysis Feature Extraction Method for Hyperspectral Images
Abstract:A new wavelet-based independent component analysis (ICA) feature extraction method is proposed according to characteristic of the hyperspectral data. In order to solve the problem that ICA is too sensitive to the noises, the proposed method makes full use of the advantages from wavelet transform for noise reduction in hyperspectral data. Then ICA is implemented into the feature extraction process for dataset before and after noise reduction and Support Vector Machine (SVM) classifier is selected for land cover classification experiment to evaluate the performances of feature extraction. Experimental results confirm the effectiveness of the proposed approach, which is good at eliminating the noise in the original image data to improve the accuracy of feature extraction and classification for hyperspectral data
Keywords: hyperspectral remote sensing feature extraction wavelet decomposition independent component analysis
基金:
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
小波降噪与独立成分分析的高光谱特征提取研究
评论
全部评论0/1000