Abstract
Abstract: :
Purpose:Defocus and aberrations degrade the retinal image of high spatial frequencies compared to low spatial frequencies, resulting in decreased contrast sensitivity at high frequencies. Georgeson and Sullivan (1976) found that observers matched the contrast of high and low spatial frequencies so that high contrast images of all frequencies appeared subjectively to have equally high contrast despite the effects of retinal blur – an effect called contrast constancy. Our previous experiments have explored the effects of refractive defocus and scatter blur on contrast matching, demonstrating contrast constancy with suprathreshold simple grating stimuli. We further explore this effect as it relates to more natural images by evaluating effects of varying retinal image size (i.e. shifting the spatial frequency spectra of the image) on contrast matching of complex, natural stimuli. Methods: 4 observers were presented with a grayscale image of a face (image 1) and a forest (image 2) at contrasts of 0.075, 0.15, 0.30, and 0.60. The perceived contrast of these images was matched with the same image subtending different angular retinal sizes (2x, 1x, 0.5x, and 0.25x standard size image). Contrast sensitivity functions were also determined. Results: While contrast sensitivity functions showed a typical decrease in sensitivity at high spatial frequencies relative to lower spatial frequencies, contrast matching functions were unaffected by the change in the spatial spectra of the images at different image sizes. Results for image 1 (face) and image 2 (trees) were similar. Conclusions: Observers demonstrate contrast constancy for suprathreshold natural stimuli despite defocus and aberrations induced by viewing images which subtend small angular retinal sizes. The amount of constancy is consistent across images of varying sizes. These results suggest that there is an adaptive mechanism responsible for maintaining constancy of image perception across varying spatial frequencies as we scan our natural environment.
Keywords: adaptation: blur • scene perception • imaging/image analysis: clinical