Related Experiment Videos
The perception of natural contour
D L Gilden1, M A Schmuckler, K Clayton
1Department of Psychology, Vanderbilt University.
Psychological Review
|July 1, 1993
Summary
Human perception of fractal patterns relies on signal-noise distinctions, not self-affinity. This suggests a natural principle of perceptual organization for viewing contours.
Area of Science:
- Visual Perception
- Cognitive Science
- Fractal Geometry
Background:
- Natural forms often exhibit fractal properties, suggesting human sensitivity to their statistical characteristics.
- Understanding how humans perceive and discriminate between fractal patterns is crucial for visual science.
Purpose of the Study:
- To investigate the perceptual protocols underlying human discrimination of fractal contours.
- To construct computational models that mimic human visual sensitivity to fractal properties.
Main Methods:
- Examined perceptual protocols for discriminating between fractals and other random contours.
- Developed algorithms with sensitivities matching human observers for fractal discrimination.
Main Results:
- Constructed algorithms that successfully discriminate fractals based on signal-noise distinctions, mirroring human performance.
- Demonstrated that self-affinity in fractals is not perceptually recovered by these algorithms or human observers.
Conclusions:
- Human perception of fractal contours relies on a signal-noise distinction rather than recovering self-affinity.
- Proposes a principle of perceptual organization where contour is naturally viewed in terms of signal and noise.