针对不同融合算法的质量评价指标性能评估

Performance evaluation of image fusion quality metrics for the quality of different fusion methods

融合质量评价指标的性能对图像融合方法的提出和改进十分重要。讨论了目前常用的三种参考图像的构建方法,并对常用融合质量评价指标(相关系数CC、偏差指数RB、结构相似度SSIM、均方根误差RMSE、光谱角SAM和ERGAS)从三方面进行性能评价分析:(1)功能相似指标间评价结果的一致性;(2)对同一融合算法、不同图像的评价结果的鲁棒性;(3)与目视评价结果的一致性。中、高空间分辨率遥感影像融合实验结果表明:CC、SSIM和SAM选择原始低分辨率多光谱图像作为参考图像性能最好,仅适合做融合图像光谱质量评价;RMSE和ERGAS选择原始低分辨率多光谱图像或高分辨率图像作为参考图像性能最好,因此这两个指标既可以做融合图像光谱质量评价,也可以做空间质量评价。

Image fusion quality metrics （IFQMs） form an essential part in the development of image fusion techniques. Focuses on the case of the reference that was missing in most cases. Six frequently used IFQMs which were correlation coefficient （CC）, relative bias （RB）, structure similarity index（SSIM）, root-mean-square error （RMSE）, spectral angle mapper （SAM） and relative dimensionless global error （ERGAS） based on different reference images are being tested to check their ability to measure the quality similarity among the images. The ability of the six IFQMs mainly in three aspects： （1） Consistency with other similar IFQMs; （2） Robustness to different testing images; （3） Consistency with the visual evaluations. Experimental results show that CC, SSIM and SAM taking the original low resolution multi-spectral image as reference will achieve the best performance. RMSE and ERGAS were robust with references which show that the two indicators can perform well both in spectral quality evaluation and spatial equality evaluation.