Related Experiment Videos
Denoising and contrast constancy
1Department of Optometry, Bradford University, Richmond Road, Bradford BD7 1DP, UK. w.h.mcilhagga@bradford.ac.uk
Vision Research
|September 11, 2004
Summary
Perceptual contrast constancy allows us to see object contrast consistently, regardless of size or spatial frequency. This study reveals perceived contrast relies on an optimal thresholding estimator, similar to wavelet denoising techniques.
Area of Science:
- Visual perception
- Computational neuroscience
- Image processing
Background:
- Contrast constancy is crucial for stable object perception.
- Visual system infers stable object properties from variable retinal images.
- Size and spatial frequency alter retinal contrast and detectability.
Purpose of the Study:
- To investigate the underlying mechanism of contrast constancy.
- To determine if perceived contrast is based on an optimal estimation process.
- To link visual perception mechanisms to image processing techniques.
Main Methods:
- Analysis of perceived contrast under varying size and spatial frequency.
- Modeling perceived contrast using optimal thresholding estimators.
- Comparison with established image processing algorithms like VisuShrink.
Main Results:
- Perceived contrast is accurately predicted by an optimal thresholding estimator.
- This estimator is mathematically identical to the VisuShrink wavelet denoising estimator.
- Demonstrates a direct link between visual perception and image denoising algorithms.
Conclusions:
- Perceived contrast is achieved through an optimal estimation process in the visual system.
- The brain employs a mechanism analogous to VisuShrink for contrast estimation.
- This finding offers insights into the computational principles of visual perception.