Laplacian-Gaussian analysis for generic representation of maneuvering affordance

On the common base of the randomness of natural texture, two universal imaging rules, perspective projection and self-similarity, are combined for detecting almost horizontal space in complex scene. Scale shift due to perspective projection is extracted through Laplacian-Gaussian image analysis. By identifying scale shift to invariant measure associated with a fractal attractor, the expansion of maneuvering area is estimated. Estimated maneuvering area is shown to be visualized on the scene image without explicit geometry of the roadway.