Abstract
This paper has clarified the difference in image noise evaluation performance between the Wiener spectrum based method and the cooperative vision model based method by comparing both methods' noise evaluation values for the identical evaluated images composed of two groups of monochrome patch images with a uniform optical density. The first group and the second group consisted of images with the identical output resolution and various types of dot structure (or texture) and with various kinds of output resolution and the same type of dot structure, respectively. Consequently the following differences were clarified:(1)for the first group, the vision model based method was considerably superior in agreement with subjective judgments to the Wiener spectrum based method;(2)for the second group, both methods agreed well with subjective judgments;(3)the vision model based method specified viewpoints where image noise was conspicuous. The above results suggest that the vision model based method covers a wider range of application than the Wiener spectrum based method.
