you won’t find a simple analytic solution by just googling it, because there remain too many unknowns. (already noted by starcrunch, but perhaps overloooked)
what you’d need to know before even beginning an analytic integration is: what are the values of those two real roots you’d be integrating between? fe, are a,b,c,d fixed constants rather than variables? if you can somehow KNOW those two roots (there might only be two, but what if 1 or 3, or none (if you allow a=0), then what??), then several approaches might work, ottomh…
1 - would you know how to reword the problem so that you’re solving for 0 instead of 2? (hint: d=d-what?)
(otoh, if not, then you might want to reconsider tackling this problem)
2 - would you then know how to find its real roots? (always this cubic? if so, hint, else hint)
3a - would you then know how to get the antiderivative (indefinite integral, hint)? if so, just integrate it between the two roots.
3b - or manually (ie iterative computation) do a finite sum where dx=1 pixel between the roots. (or approximate the curve with a polygon generated from those same dx=1 vertices and calculate its signed area - it’ll be more computation overall, but perhaps conceptually simpler?)