Yeah, they do look kind of like “bubbles” rather than depressions in that photo. I think part of it is they seem to be shallower craters than we often see too, either they are older and more weathered, or maybe it is just a trick of the resolution.
Well, I think it’s a matter of what we’re used to seeing. Normally protrusions on vertical surfaces like walls will have a highlight on their upper surface (and shadow on their lower side) because they’re typically lit from above whether by the sun or by artificial lighting. So when our mind sees craters with highlights on top it makes the natural assumption based on experience that they actually protrude. When you rotate the image so the highlight isn’t on top, that experiential logic is broken and you’re more likely to see the crater as the depression that it actually is.