Real-Time Point-Based Global Illumination
aduprat.com
aduprat.com
The approach would be just exhaustive summation, I suppose. Would symbolic integration be equally expansive for simple enough geometry? Simplified approximations to the geometry, i.e. bounding boxes, would probably not look real enough. I get, there are all kinds of approaches.
I heard Monte Carlo method used to limit the search space to a random subset.
A lot of work has been done on using analytic integration techniques, as well as higher quality numeric approximations, for global illumination, in particular with radiosity methods. https://en.m.wikipedia.org/wiki/Radiosity_(computer_graphics...
http://www.mi.uni-koeln.de/c/mirror/www.cs.curtin.edu.au/uni...
Imperfect Ray-Bundle Tracing for Interactive Multi-Bounce Global Illumination
https://www.youtube.com/watch?v=VHbHOQ1NRuw
And so did Unity:
1. The color is quantified into evenly spaced segments (ie a 3 dimensional image)
2. Because it is a 3 dimensional image the GPU's filtering hardware can do efficient lookups over a volume using mip mapping techniques.
The approach was originally described in a siggraph paper and there have been a lot of variations (which is usually another sign that something works well). Someone could create an implementation on their own, it isn't much of a stretch. You just rasterize into a voxel image then do a few cone traces at every pixel you shade, guided by the BDRF (a good one to look at would be ggx).