Hi.. I'm developing a small project, where I have a process running which renders in OpenGL a 3d world - simple geometry and solid colors. From time to time a screenshot of the viewport is taken and sent via UDP socket to another process wherelse which is an AI agent. The agent receives the image, find shapes, objects and sends a response through the same socket, to drive a little car through the world. Both programs and going well.. but I'd like a bit of ideias how to parse that image. The project is obviously being developed in C++, but any documentation in any language would be apreciated.
So any I'm simply asking some info (decent one preferably) on algorithms, methods to recognise stuff on images. No, not code!
Second, how do I take that screenshot from the viewport in OpenGL ?!? I' using glut to ease up a bit.
Thank you for your attention.