Multi-scale filtering of image feature associated with maneuvering affordance

Scale information is mapped onto an intentional coordinate system for detection and visualization of maneuvering affordance in complex scene. The brightness distribution of the directional scale image is identified with fractal attractor to implement a pointwise filters for reducing existing background noise. For this purpose, the capturing probability for the attractor is generated and dominant scale information of roadway texture is emphasized. By applying Laplacian mask of the capturing probability, the expansion of the affordance is visualized in the scene image. The detectability of affordance patterns has been verified through experimental studies.