农业机械学报2014，Vol.45Issue(11)：264-271,8.DOI:10.6041/j.issn.1000-1298.2014.11.041
基于LBV变换与小波变换的OLI图像融合方法
Fusion of OLI Image Based on LBV Transform and Wavelet Transform
摘要
Abstract
The aim of this study is to seek out the most suitable image fusion algorithm for OLI image of Landsat 8 satellite acquired in June 9,2010,taking Yuyang country in Shaanxi Province as study area.Five kinds of image fusion algorithms have been employed,which are Brovey transform,High-pass filter transform,HIS transform,PCA transform and LBV-wavelet RF.The effectiveness of the five fusion algorithms has been evaluated based on spectral fidelity,high spatial frequency information gain,and supervised classification accuracy.Firstly,by visual evaluation this study evaluated whether fused images preserved spectral information of original multispectral image well,and whether retained texture and edges information of panchromatic image and avoided texture blurring.Secondly,by quantitative evaluation,spectrum character of fused images was analyzed by using gray average difference and gray root mean square error.Integration of the high frequency detail information of panchromatic images to fused images was analyzed by using correlation coefficient average and correlation root mean square error.The supervised classification accuracy of fused images was evaluated by using Kappa coefficient and overall classification accuracy.Results showed that LBV-wavelet RF was the best method in retaining spectral information of original multispectral image,and not causing spectral distortion,as well as achieving the highest SVM supervised classification accuracy.Overall classification accuracy and Kappa coefficient of fused image using this method were 84.01％ and 0.787,achieved noticeable growth of 13.45％ and 15.91％ than original multispectral image.The proposed OLI image fusion algorithm could provide far more detailed topographic information compared with original multispectral dates and better service for improving visual interpretation and supervised classification accuracy.关键词
黄土高原水蚀风蚀交错带/OLI图像/融合算法/小波分解/LBV变换Key words
Water-wind erosion crisscross region of the Loess Plateau / OLI image / Fusion algorithm/Wavelet decomposition /LBV transform分类
信息技术与安全科学引用本文复制引用
刘炜,王聪华,杨晓波,雒伟群..基于LBV变换与小波变换的OLI图像融合方法[J].农业机械学报,2014,45(11):264-271,8.基金项目
国家自然科学基金资助项目(41361044、61162025)和西藏民族学院青年学人培育计划资助项目(13myQP09) （41361044、61162025）
