基于中巴资源卫星的多源遥感影像融合研究

Multi-source Remote Sensing Image Fusion Research Based on the China-Brazil Resource Satellite

随着遥感获取数据手段的日益增多,遥感影像的融合技术备受关注。针对资源二号多光谱影像光谱信息丰富、分辨率低,资源二号全色影像分辨率高、纹理信息丰富的特点,以河南省郑州市的资源二号多光谱影像与全色影像为例,分别采用基于主成分分析、小波变换、比值变换和高通滤波法等4种融合方法进行融合实验,并对融合后的影像在均值、标准方差、偏差指数以及光谱保真度等4个方面进行了对比,探讨了资源二号多光谱影像与全色影像融合的方法和效果。

With the growing of remote sensing data access methods, remote sensing image fusion technology has been concerned. The image of CBERS 2 has abundant information and low resolution, but the panchromatic image has high resolution and rich texture information. Taking the CBERS 2 multi-spectral image and panchromatic image of Zhengzhou City, Henan Province, as example, fusion experiments were carried out using four kinds of fusion methods including principal component analysis （PCA）, wavelet transform （WPC）, the brovey transformation（Brovey） and high -pass filtering（HPF） method. Comparison of the fusion image was made in the mean, the standard deviation, the deviation index, and the spectral fidelity, and the multi-spectral image and panchromatic image fusion method and effect of CBERS 2 was studied.