中文摘要：

采用Brovey、HIS、PanSharpen、Gram—Schmid、PCA、Wavelet等融合方法对World—View全色与Geoeye—1多光谱影像进行融合处理，在传统的光谱还原、纹理清晰度指标的基础上，引入植被权重的改进相似度定量评价法，其以人眼的视觉特性为出发点，侧重于图像的结构信息和光谱与纹理的均衡度，并着重考虑林业遥感尤其城市森林资源监测应用中对突出植被信息的要求，利用MatLab构建评价程式，通过对G—MMSIM及常规指标的综合比对分析，认为此法与目视判读较为接近，可以在评价实践中予以应用，同时，得到PCA和Gram—Schmid效果最优，PanSharpen稍差的评价结果。

英文摘要：

The remote sensing data of WorldView - 1 and GeoEye - 1 were fused with six kinds of methods including Brovey, HIS, PanSharpen, Gram-Schmid, PCA, and Wavelet. Based on the traditional spectrum unfolding method and texture resolution index, the evaluation method of improved similarity of vegetation weight was introduced in the paper which considers the characteristics of human visual system, in particular the structural information of images and evenness of spectrum and texture, as well as the requirements on vegetation information extraction in the applications of forest remote sensing and urban forest resource monitoring. Then, an evaluation model was constructed using Matlab. Through the comparison analysis of G_ MMSIM and regular indexes, it is believed that the method shows good agreement with visual interpretation. Meanwhile, the above analysis indicated that the fusion methods of PCA and Gram-Schmid are the best ones, followed by PanSharoen.