Paper data
Title:
Structure Sensitive Sampling of Noisy Fractal Patterns Author(s): Kamejima Kohji, Osaka Inst. Tech. Page numbers in the proceedings: Volume III pp 547-550 Session: Object Recognition
Paper abstract
An adaptive sampling scheme is presented for detecting complex patterns in noisy imagery. By representing imaging processes in terms of unknown contraction mappings, capturing probability is generated for fractal attractors approximating observed patterns. The totality of local maxima of the capturing probability is shown to yield a pattern sensitive sampling of the fractal attractor. The detectability of the sampling scheme has been verified through simulation studies.
Paper
|