How do you capture multiple illustrations?

[Lech Rzedzicki <xchaotic@gmail.com>]

Hi.

We want to capture groups of illustration using DB5 to represent an
illustrated book content.
For the individual illustration we will of course use <mediaobject>
but we're not sure how to group them nicely.
<figure> is sometimes ok, but other times the illustrations have
different caption, no captions or one illustration can be a "child" of
another (for instance it shows a zoomed in portion of the first
illustration).
What Docbook elements would you recommend for consistently
representing such varied scenarios?
Using section or a para to group them is a possible solution, but
doesn't introduce much semantic value into the markup and doesn't seem
very elegant to me.
Ideally I would want to reuse something that's already found in
Docbook Publishers SC, if not then from the full Docbook5, failing
that, we might have to come up with some custom elements or
attributes?


I really appreciate any feedback,

Lech