I'm sure they would, those being the people who are coming to describe almost anything achieved by computer processing as "AI", even if 'intelligence'/'intellect are not really involved at all.
I'm pretty sure that the sort of image processing we're talking about could be fairly easily achieved by a computer executing a traditional algorithm-based program ('imperative' programming) written by a human being, such that the computer was doing nothing other than strictly, but totally 'dumbly', obeying a set of explicit instructions, with absolutely no 'machine intelligence' involved at all.
In terms of 'true AI', I imagine that there is probably a lot of debate as to what we actually mean by, and should regard as, "intelligence". I remember many moons ago reading an article that was saying that advances in computer capabilities were turning some traditional thinking on it's head. It pointed out that, traditionally, virtually no-one would disagree that one needed a very high level of 'intelligence' to be a Chess Grand Master, nor would they disagree that, say, recognising a human face was a trivial exercise, which even someone of 'minimal intelligence' could do. However, those traditional views needed to be re-considered when computers became able to beat the Grand Masters at chess, at a time when computer facial recognition was not really good enough to be useful.
For what it's worth, my personal view would be that, as a starting point, 'true AI' has to be able to do things beyond merely executing explicitly programmed instructions (which I imagine is all the chess-playing computers of the time were doing).
