用于耕地信息提取的多光谱与全色遥感影像融合方法研究

Study on fusion method of multi-spectral and panchromatic remote sensing image for farmland information extraction

低分辨率多光谱图像和高分辨率全色图像通过图像融合,可最大限度地发挥两类图像数据的技术优势,达到信息互补和充分表达地物信息的目的。首先,采用IHS变换、PCA变换对某一地区耕地地物信息的多源遥感图像进行融合,分析传统融合方法的不足,之后,以提高光谱信息和空间信息的准确性为目标,采用小波变换与IHS变换结合的方法,最终得到兼具较好空间分辨率和耕地光谱信息的融合图像。通过主观和客观融合质量评价,确认小波与IHS变换相结合的方法优于传统融合方法,该融合方法可为宁夏平原耕地信息的遥感影像融合提供参考借鉴。

Fusion image with low resolution multispectral image and high resolution panchromatic,can maximize to show technology advantages of the two kinds of image data,and can achieve the goals of information complementation and ground information expressing adequately.The IHS and PCA transform was adopted to fuse multisource remote sensing image of farmland feature information,and then,shortages of the traditional fusion method were analyzed.To improve the accuracy of spectral information and spatial information for the target,the better spatial resolution and spectral information of cultivated land fusion image was obtained through the wavelet transform combined with IHS transform method eventually.Through the subjective and objective fusion quality evaluation,method of the wavelet transform combined with IHS transform was superior to traditional fusion method.Besides,this fusion method can provide a reference to image fusion of cultivated land multisource remote sensing in Ningxia plains.