**2. Materials and Methods**

#### *2.1. Image Datasets*

Table 1 shows the two image datasets used for the experiments. The first was collected in May 2008 and covers the city of Nihonmatsu, Japan. The second was collected in May 2006 and covers the city of Yokohama, Japan. The two IKONOS images datasets were provided by the Japan Space Imaging Corporation, Japan. The spatial resolutions of the PAN and the MS images in these datasets were 1 m and 4 m, respectively. The original dataset contained PAN images with 1024 × 1024 pixels and MS images with 256 × 256 pixels for Nihonmatsu, and PAN images with 1792 × 1792 pixels and MS images with 448 × 448 pixels for Yokohama.


**Table 1.** Characteristics of the original, training, and test images of the image datasets. MS: multispectral, PAN: panchromatic.

The training image datasets had high-resolution PAN images with 256 × 256 pixels for Nihonmatsu, and high-resolution PAN images with 448 × 448 pixels for the Yokohama.

To evaluate the quality of the PS images, we experimented with the test images and original images according to the Wald protocol [36]. The test images were used to evaluate the numerical image quality, and the original images were used as reference images for numerical and visual evaluation. We regarded the original images as ground truth images. The spatial resolution of the test PAN images was reduced from 1 to 4 m and that of the test MS images was reduced from 4 to 16 m. Hence, the test image datasets had PAN images with 256 × 256 pixels and MS images with 64 × 64 pixels for the Nihonmatsu, and PAN images with 448 × 448 pixels and MS images with 112 × 112 pixels for the Yokohama. The training and test images were downsampled images of the original image with bicubic spline interpolation.
