Proc. SPIE. 7240, Human Vision and Electronic Imaging XIV
KEYWORDS: Edge detection, Digital photography, Detection and tracking algorithms, Denoising, Image analysis, Image quality, Digital imaging, Image enhancement, Modulation transfer functions, Algorithm development
Motivated by the reported increase in sharpness by image noise, we investigated how noise affects sharpness perception.
We first used natural images of tree bark with different amounts of noise to see whether noise enhances sharpness.
Although the result showed sharpness decreased as noise amount increased, some observers seemed to perceive more
sharpness with increasing noise, while the others did not. We next used 1D and 2D uni-frequency patterns as stimuli in
an attempt to reduce such variability in the judgment. The result showed, for higher frequency stimuli, sharpness
decreased as the noise amount increased, while sharpness of the lower frequency stimuli increased at a certain noise level.
From this result, we thought image noise might reduce sharpness at edges, but be able to improve sharpness of lower
frequency component or texture in image. To prove this prediction, we experimented again with the natural image used
in the first experiment. Stimuli were made by applying noise separately to edge or to texture part of the image. The result
showed noise, when added to edge region, only decreased sharpness, whereas when added to texture, could improve
sharpness. We think it is the interaction between noise and texture that sharpens image.
To overcome shortcomings of digital image, or to reproduce grain of traditional silver halide photographs, some
photographers add noise (grain) to digital image. In an effort to find a factor of preferable noise, we analyzed how a
professional photographer introduces noise into B&W digital images and found two noticeable characteristics: 1) there is
more noise in mid-tones, gradually decreasing in highlights and shadows toward the ends of tonal range, and 2)
histograms in highlights are skewed toward shadows and vice versa, while almost symmetrical in mid-tones. Next, we
examined whether the professional's noise could be reproduced. The symmetrical histograms were approximated by
Gaussian distribution and skewed ones by chi-square distribution. The images on which the noise was reproduced were
judged by the professional himself to be satisfactory enough. As the professional said he added the noise so that "it
looked like the grain of B&W gelatin silver photographs," we compared the two kinds of noise and found they have in
common: 1) more noise in mid-tones but almost none in brightest highlights and deepest shadows, and 2) asymmetrical
histograms in highlights and shadows. We think these common characteristics might be one condition for "good" noise.