A reliable quality assessment procedure for pansharpening methods is of critical importance for the development of the related solutions. Unfortunately, the lack of ground-truths to be used as guidance for an objective evaluation has pushed the community to resort to either reference-based reduced-resolution indexes or to no-reference subjective quality indexes that can be applied on full-resolution datasets. In particular, the reference-based approach leverages on Wald's protocol, a resolution degradation process that allows one to synthesize data with related ground truth. Both solutions, however, present critical shortcomings that we aim to mitigate in this work by means of an alternative no-reference full-resolution framework. On one side we introduce a protocol, namely the reprojection protocol, which allows to handle the spectral fidelity problem. On the other side, a new index of the spatial consistency between the pansharpened image and the panchromatic band at full resolution is proposed. The experimental results show the effectiveness of the proposed approach which is confirmed also by visual inspection.
翻译:不幸的是,缺乏地面真相作为客观评价的指导,促使社区诉诸基于参考的低分辨率指数或可适用于完整分辨率数据集的不参考的主观质量指数,特别是,基于参考的方法对Wald的规程产生了杠杆作用,这是一个分辨率退化过程,使得人们能够将数据与相关的地面真相结合起来。但是,这两种解决办法都存在严重缺陷,我们力求通过一个替代的不参考的全面分辨率框架在这项工作中减轻这些缺陷。一方面,我们引入了一个议定书,即可处理光谱忠诚问题的重新预测程序。另一方面,提出了一个全分辨率图像和全分辨率板段之间的空间一致性新指数。实验结果显示了拟议方法的有效性,也得到了视觉检查的证实。