I am trying to make a first person videogame in OpenGL. The principal idea is to make models in blender and parse this models to OpenGL (.obj format).

Once done, I start to work with the collisions, but I don´t know how to do certain collisions because they are complex, like a house (I need to detect collisions of the walls but the player can enter the house through the door).

The idea is to automate this process as much as possible.

My first idea was add in blender cubes (as bounding-boxes) and treat them differently in the parser, but it’s very tedious and not automatic as I want.