Li Hui, Jing Linhai, Tang Yunwei
Key Laborary of Digital Earth Sciences, Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, Beijing 100094, China.
Sensors (Basel). 2017 Jan 5;17(1):89. doi: 10.3390/s17010089.
Since WorldView-2 (WV-2) images are widely used in various fields, there is a high demand for the use of high-quality pansharpened WV-2 images for different application purposes. With respect to the novelty of the WV-2 multispectral (MS) and panchromatic (PAN) bands, the performances of eight state-of-art pan-sharpening methods for WV-2 imagery including six datasets from three WV-2 scenes were assessed in this study using both quality indices and information indices, along with visual inspection. The normalized difference vegetation index, normalized difference water index, and morphological building index, which are widely used in applications related to land cover classification, the extraction of vegetation areas, buildings, and water bodies, were employed in this work to evaluate the performance of different pansharpening methods in terms of information presentation ability. The experimental results show that the Haze- and Ratio-based, adaptive Gram-Schmidt, Generalized Laplacian pyramids (GLP) methods using enhanced spectral distortion minimal model and enhanced context-based decision model methods are good choices for producing fused WV-2 images used for image interpretation and the extraction of urban buildings. The two GLP-based methods are better choices than the other methods, if the fused images will be used for applications related to vegetation and water-bodies.
由于WorldView-2（WV-2）影像在各个领域得到广泛应用，因此对于高质量的WV-2融合影像以用于不同应用目的的需求很高。鉴于WV-2多光谱（MS）和全色（PAN）波段的独特性，本研究使用质量指标和信息指标，并结合目视检查，评估了八种用于WV-2影像的先进融合方法的性能，这些方法涉及来自三个WV-2场景的六个数据集。本研究采用了归一化植被指数、归一化水体指数和形态学建筑物指数，这些指数在土地覆盖分类、植被区域提取、建筑物和水体提取等相关应用中广泛使用，以评估不同融合方法在信息呈现能力方面的性能。实验结果表明，基于薄雾和比值法、自适应Gram-Schmidt法、使用增强光谱失真最小模型的广义拉普拉斯金字塔（GLP）法以及基于上下文增强决策模型法，是生成用于图像解译和城市建筑物提取的融合WV-2影像的良好选择。如果融合影像将用于与植被和水体相关的应用，那么两种基于GLP的方法比其他方法是更好的选择。
