It isn't clear whether the question is specifically about pictures, and if so, what kind of pictures. Under one interpretation, English; under another interpretation, probably no language. The differences that I am alluding to are about the perspective that a person might take given a certain task, and how those perspectives relate to grammar. I suggest that attention should be paid to experimentally controlling the assumptions made by language subjects, as is standard in semantic and pragmatic fieldwork.
The first question is, what is "normal behavior" for reporting events as you observe them. It is not normal to narrate a thing happening in front of you (plural), instead you might make comments to the guy next to you, such as "OMG he just kicked the dude!" or "He's gonna fall on his butt" or "He's beating the pants off of Tom". That is one narrative context, which doesn't involve pictures. A similar narrative context is when you are reporting events in real time to a person who can't see what is going on. A third context is professional broadcasting, which may or may not be the same (because professional broadcasters may follow a professional style that is not used by ordinary people).
There are various theories of information about events related to "when" questions, for example the Neo-Reichenbachian theory which seems to be the most widely adopted in linguistics, which relies on three time concept: utterance time, topic time and event time. Utterance time and event time are fairly straightforward (it's about external actions, namely the event and the act of uttering). Topic time is, to put it one way, "the time which the sentence is about", which is not totally self evident. In telling a story, the narrator may "put themselves in the past" and therefore use a present tense form, or they may "remain in the present" and talk about past events as though they are in the past.
This flexibility of perspective seems to exist in every language, but nevertheless there are cultural conventions. The present tense was used in the Star Trek "Darmok" narrative, even though the events were in the (far) past.
A picture being a representation of an event, that abstraction can be temporally removed from the event that it represents, but w.r.t. "the things that happened", it is (especially if it is a photograph) less-removed from the event than a verbal description or stick-figure drawing – it is more vivid, which encourages a "current time" perspective. Therefore, you can easily describe what's going on in the picture in the present, or in the past.
However, the perspective that you take is highly influenced by why you are saying anything in the first place. If somebody asks a "present-relevant" question about the question such as "which person is Julius Caesar", the question is not about what JC was doing 2 millennia ago, it is about the picture that we see right now, so the natural response would be "He is the guy wearing red velvet". If the question is what event this picture reports, you could say "JC executed his enemy" (assuming a fait accompli in the picture) and "JC is executing his enemy" would be unnatural. However, if the picture depicts the actual execution, that would be a natural response.
I do not believe that there are languages that have special grammatical rules for pictures. Instead, there may be cultural conventions governing the perspective that you use in talking about an event. Photographs especially have a special ability to make past events more "vital" and present-relevant.