In our previous paper, we investigated the effects of image noise on the perception of image sharpness using white noise, and one- and two-dimensional (1D and 2D, respectively) single-frequency sinusoidal patterns as stimuli. Our results showed that with high-frequency stimuli, sharpness decreased as noise level increased; whereas, with low-frequency stimuli, sharpness reached a maximum value at a certain level. The sharpness of the rectangular grating and the checkerboard pattern, both of which have clear edges, decreases more drastically even than the sinusoidal wave with the highest frequency. This paper builds on our previous study by evaluating color natural images rather than black-and-white, single-frequency patterns. Our results indicate that the effects of noise are more remarkable in images of objects containing more low-frequency components, in blurred rather than sharp images, and in strongly blurred images. The dependence of sharpness on frequency, as found in this study, is consistent with results from the preceding study; the results we obtained using black-and-white, single-frequency patterns were verified using color images of natural objects. Furthermore, we acquired new information on the influence of texture, which we did not investigate in our previous studies, i.e., noise has a negligible effect on images with a large number of edges; whereas, we observed the most significant effects for images without such a texture.

View full abstract