Anticipative generation and in-situ adaptation of maneuvering affordance in naturally complex scene

A fractal representation of the maneuvering affordance has been introduced for anticipative decision making through satellite-vehicle-roadway network. Based on the randomness ineluctably distributed in naturally complex scenes, a probability for capturing not-yet-identified fractal attractor is estimated and applied to anticipative generation of roadway models. Through anticipative adaptation, the fractal code is transferable for cooperative road following process where the boundary of open spaces is defined through in-situ adaptation of the maneuvering affordance to observed scenes. The coding and adaptation scheme was applied to typical scenes to demonstrate the randomness-based fractal coding.