Poisson-Boltzmann is not sufficient to determine surface charge density, you need something more. For example, this may be the surface potential, if you can measure it as in the case of metal surfaces. If it's not a metal, then you need an adsorption model (e.g., Davies isotherm or something more complex, such as the full ion-surface interaction potential).
@Jochen Potential can be used to compute the electric field near the surface and then using the boundary conditions one can find the the surface charge density.
@Radmir I am not considering ions in my solvent, so I will just need the solution for uniform dielectric with finite dielectric constant.
Assume a simple case, say you have a cavity with certain charges. Then the induced charges are due to reaction field potential. Now one can in principle compute the potential due to the original charges in the cavity and the induced charges (linear superposition principle) and obtain the potential at a point outside the cavity. Now there are softwares which compute the potential using methods like BEM. So one can solve the inverse problem to deteremine the charge on the surface.
I was wondering if has already been develpoed by someone or if more elegant techniques are out there.
NOTE that I am not interested in higher moments of water or anything. It is plain simple continuum linear resposne electrostatic problem.
If the cavity is spherical, then you can use the respective Green function. I.e., you only need the solution for a point charge somewhere in the cavity, and then sum or integrate over the charges in your cavity.
If it's not spherical, then you can use Comsol Multiphysics. I think they offer a 1 month trail version; this will be enough to try it. It is also possible with Mathematica/Maple/MatLab, but it will be harder.