If p were to be on the lower side of the image, I think f() function will be flipped because the upper region will have more difference in intensity compared to the lower portion. Thus, gxf will also be flipped the same way as g stays unchanged.
If p were to be on the lower side of the image, I think f() function will be flipped because the upper region will have more difference in intensity compared to the lower portion. Thus, gxf will also be flipped the same way as g stays unchanged.