论文部分内容阅读
采用影像融合的方法,分析中国自行研发的资源一号02C卫星数据在森林资源分类中的应用效果。以2013年浙江省龙泉地区的多光谱和全色影像为数据源,在影像配准、校正基础上,经Brovey、Gram-Schmidt、高通滤波(HPF)、HSV、IHS、主成分与小波变换等7种融合方法,采用定性与奇异值分解的定量分析评价其影像的融合效果。结果表明:HSV、IHS和HPF变换为较适宜该影像数据的融合方法;进一步通过分类后的精度评价及验证可知,HPF融合方法最优且该融合后的影像适宜采用最大似然法进行分类。
Using the method of image fusion, this paper analyzes the application effect of resource No.1 02C satellite data developed by China in forest resource classification. Based on the multi-spectral and panchromatic images in Longquan area of Zhejiang province in 2013, based on the registration and calibration of images, the data were analyzed by Brovey, Gram-Schmidt, HPF, HSV, IHS, principal component and wavelet transform Seven kinds of fusion methods were used to evaluate the fusion effect of the images by quantitative analysis of qualitative and singular value decomposition. The results show that the HSV, IHS and HPF transform are more suitable for the fusion of the image data. After the classification accuracy evaluation and verification, it is found that the HPF fusion method is the best and the fused image is suitable for classification by the maximum likelihood method.