Neuro-fuzzy posture estimation for visual vehicle guidance

This paper presents a neuro-fuzzy approach to visual guidance of a mobile robot vehicle in local maneuvers. It is based on the transfer of the skills of an experienced driver to an automatic controller. The resulting controller processes video sensor data to generate corresponding steering and velocity commands in real time. Neither a geometric environment model nor analytic models of the video sensor or the vehicle kinematics are required. In contrast to previous work of the authors, the controller commands are generated by a two-stage processing structure. A first stage estimates the vehicle posture relative to the desired goal based on the video images. A guidance controller uses the estimated posture to calculate appropriate commands in a second stage. The approach is exemplified and validated by the design and implementation of a visual parking controller for the experimental robot vehicle MACROBE.