How to extract and to exploit vision data for autonomous and mobile robots to operate in known environments