Well in the same way that Epicurus held the swerve to exist without suggesting a mechanism for it, I would think that he would presume that there is a method of storing pictures in the mind, even if he dd not suggest a mechanism. I believe in my own introspection that I can picture in my mind a picture of the Mona Lisa, for example, and I can call from memory what seems to be a "picture" of it, which I see no reason not to consider as related to storing "images."
Maybe we are thinking of images differently, in that I see them as simply variation of the particles that are also seen and heard, not as something fundamentally different that would require a different storage mechanism, although certainly that's a possible reading of what it going on here.
My main issue is that I think that there is clearly a mechanism of memory for things we see, we hear, and touch, and taste, and I don't see any reason to conclude that memory of those sensations should be presumed to be consumed by this mechanism of the effect of images on the mind, which seems to me to be a separate category of phenomena on its own. Having some kind of general memory function is so basic to human operation that it would be impossible either to write the poem or read it and get any understanding of it unless we had a fundamental capacity to remember things that happen to us, so it seems to me to be highly unlikely that such a fundamental capacity would escape mention until the latter part of book four in these passages.
If that were what was meant by this aspect of Epicurus theory I would think the mechanism would be much more prominent in other parts of the texts
(Oops I missed seeing Don's comments, let me address those too)
There's also the Sedley idea of one sending images *to* the gods which sounds like we can create images in our minds.
I had not thought of this but yes I can see that being a possible explanation of the "to" issue. In that context I was presuming that we were referring the particles that make up the images that float through the air as something that the gods were accessing directly to replenish their own particles, kind of like sucking static electricity out of the air, so it wouldn't make any difference to them what images they used to replenish themselves. But you are right that that "to" indicates something else.
And to add to this mix of confusion in my mind I can't help but this this could be related to the process of forming anticipations, but on the other hand I remember from the first time I read DeWitt's book (i would have to now look for the reference) that DeWitt didn't think that Lucretius addressed anticipations anywhere at all in the poem, so apparently DeWitt himself didn't relate this section to anticipations.
Of course all of us are capable of being wrong so any or all of my current thinking on this could be incorrect.
Most of this and probably all of it don't strike me as particularly urgent to resolve quickly, but I would like to make some progress on the issue of determining how likely it is that the operations being discussed here constitute a complete description of human memory and/or the operation of the mind in general. I think whether we are just trying to reconstruct Epicurus or whether we are stating our own views based on all sources of information up through today, we ought to be able to articulate a description of what memory seems to us to be, and I think this discussion is showing that we have some basic gaps in that description as to what memory appears to be. Are we remembering pictures, or not? And what is the relation to "images" in this part of the discussion to "pictures"? Is there any at all? - perhaps "pictures" is applicable only to what the eyes process, and "images" in this context is a deceptive word for this use in English.
So at this point I am not satisfied either with our articulation of what memory is, or of what Epicurus is saying, much less how the two come together.