**Appendix A**

**Figure A1.** Result and comparison of the proposed Pan sharpening method for the F1 (zoomed) area, which is represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.

**Figure A2.** Result and comparison of the proposed Pan sharpening method for the F5 area (zoomed), which is represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.

**Figure A3.** Result and comparison of the proposed Pan sharpening method for the F6 area (zoomed), which is represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.

**Figure A4.** Result and comparison of the proposed Pan sharpening method for the F6 area (zoomed), which is represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.


**Table A1.** Numeric results of spectral quality metrics of the Pan-sharpened images produced by selected algorithms for rural test sites (blue: highest accuracy; red: lowest accuracy).

**Table A2.** Numeric results of spatial quality metrics of the Pan-sharpened images produced by select algorithms for the rural test sites (blue: highest accuracy; red: lowest accuracy).


**Figure A5.** Result and comparison of the proposed Pan sharpening method for the F2 test frame, zoomed areas, which are represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.

**Figure A6.** Result and comparison of the proposed Pan sharpening method for the F4 test frame, zoomed areas, which are represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.

**Figure A7.** Result and comparison of the proposed Pan sharpening method for the F9 test frame, zoomed areas, which are represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.


**Table A3.** Numeric results of spectral quality metrics of the Pan-sharpened images produced by select algorithms for urban test sites (blue: highest accuracy; red: lowest accuracy).

**Table A4.** Numeric results of spatial quality metrics of the Pan-sharpened images produced by select algorithms for urban test sites (blue: highest accuracy; red: lowest accuracy).


**Figure A8.** Result and comparison of the proposed Pan sharpening method for the F3 test frame, zoomed areas, which are represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.

**Figure A9.** Result and comparison of the proposed Pan sharpening method for the F7 test frame, zoomed areas, which are represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.

**Figure A10.** Result and comparison of the proposed Pan sharpening method for the F8 test frame, zoomed areas, which are represented in a true color (RGB) combination. (**a**) MS; (**b**) Pan; (**c**) Lab; (**d**) GIHS; (**e**) GS; (**f**) HCS; (**g**) IHS; (**h**) NNDiffuse; (**i**) Ehlers.


**Table A5.** Numeric results of spectral quality metrics of the Pan-sharpened images produced by select algorithms for suburban test sites.

**Table A6.** Numeric results of spatial quality metrics of the Pan-sharpened images produced by select algorithms for suburban test sites.



**Table A7.** Band-by-band calculation results of spectral quality metrics belonging to test site F3.

### **List A1. Definition of Terms in Table 2**

**RMSE:** MN is the image size, PS(*i*, *j*) and MS(*i*, *j*) represent pixel digital number (DN) at (*i*, *j*) 'th position of Pan-sharpened and MS image.

**ERGAS:** *dh dl* represents the ratio between the pixel size of high resolution and low resolution images; e.g., <sup>1</sup> <sup>4</sup> for Pléiades data, and *n* number of bands. The RMSE represents root mean square error of band *i*.

**SAM:** The spectral vector V = {V1, V2, ... , Vn} stands for reference MS image pixels and Vˆ = {Vˆ 1, Vˆ 2, ... , Vˆ n} stands for Pan-sharpened image pixels rep reference and both have L components. **RASE:** The μ represnts the mean of *b*th band; *b* is the number of bands and RMSE represents root mean square error.

**PSNR:** The L represents the number of gray levels in the image; MN is the image size, Ir(i, j) is pixel value of reference image and Ip(i, j) is the pixel value of Pan-sharpened image. A higher PSNR value indicates more similarity between the reference MS and Pan-sharpened images.

**QAVG:** The *x* and *y* are the means of reference and Pan-sharpened images, respectively; σ*xy* is the covariance and σ<sup>2</sup> *<sup>x</sup>* and σ<sup>2</sup> *<sup>y</sup>* are variances. As QI can only be applied to one band, the average value of three or more bands (QAVG) is used for calculating a global spectral quality index for multi-bands images. QI values range between −1 and 1. A higher value indicates more similarity between reference and Pan-sharpened image.

**SSIM:** The μ stands for mean, σ stands for standard deviation; *Ir* and *Ip* represent reference and Pan-sharpened image respectively. The C1 and C2 are two necessary constants to avoid the index from a division by zero. These constants depend on the dynamic range of the pixel values. A higher value of the measured index shows the better quality of Pan-sharpened algorithm.

**CC:** Cr,f is the cross-correlation between reference and fused images, while Cr and Cf are the correlation coefficients belonging to reference and fused images respectively.

**SRMSE:** Edge magnitude (M) is calculated via spectral distance of horizontal and vertical (*Mx* and *My*) edge intensities.

**Sp-ERGAS:** *dh dl* represents the ratio between the pixel size of MS and Pan images, and *n* is the number of bands. Spatial RMSE is represented as below:

$$\text{Spatial RMSE} = \sqrt{\frac{1}{\text{MN}} \sum\_{i=1}^{M} \sum\_{j=1}^{N} \left( \text{Pan}(i, j) - \text{PS}(i, j) \right)^2} \tag{A1}$$

where MN is the image size, PS(*i*, *j*) and Pan(*i*, *j*) represents the pixel digital number (DN) at (*i*, *j*) 'th position of Pan-sharpened and Pan image.
