用基于IHS变换的SPOT-5遥感图像融合进行作物识别

    Intensity-hue-saturation model based image fusion of SPOT-5 HRG1 data for crop identification

    • 摘要: 遥感图像融合可以发挥多源遥感数据的优势。但是由于遥感数据和融合模型的多样性,目前仍难以找到一种适合于各种类型数据之间、各种应用需要的“万能”的融合算法,而是根据特定图像,特定地表覆盖状况和特定应用的需要选择适合的融合模型。SPOT-5图像是一种较新的高空间分辨率遥感图像,目前已用于运行化农情遥感监测,以弥补传统低空间分辨率遥感图像应用的不足。该文将SPOT-5多光谱和超模式全色图像进行融合,以进行中国东北地区大豆识别。对实验数据分别做基于IHS变换和PCA变换的融合处理,通过比较得出,基于IHS变换的融合

       

      Abstract: Image fusion can merge multi-source RS data outputting a better quality image. But the fusion model is dependent on the specific image type used, the specific temporal properties of the images, the specific land cover of the study area and what specific information to be extracted from the source images. SPOT-5 HRG1 image is a kind of new RS image at high spatial resolution that has been used in agricultural condition monitoring. In this study, SPOT-5 HRG1 multispectral and super mode panchromatic images were merged for soybean identification in Northeast China. Image fusions based on intensity-hue-saturation transformation (IHS) and Principal Component Analysis (PCA) were respectively done. The visual inspection and quantitative comparison of the two fusion images indicate that IHS model based image fusion of SPOT-5 HRG1 images was better for soybean identification than that of PCA model based.

       

    /

    返回文章
    返回