In order to clarify the visual information processing mechanism at each fixation point, the subject's viewing time while reading Japanese text was controlled by inserting a masking pattern in his (her) focusing area. The task was to report whether and incorrect character was included in his (her) fixation area. Two kinds of masks were used, a random dot pattern and a pattern with the correct character. The results shows that for acquiring visual information is 66 msec was sufficient and 33 msec insufficient, which was consistent with both the results of the alphabet experiment an our counting tasks. The ratio of the correct responses during random dot masking was higher than during correct character masking.

View full abstract

CCIR Report 1082(MOD F) points out that the conceptual differences between the meanings of the quality scale descriptors are not necessarily uniform. It describes the results of studies carried out in USA. Germany, France, and Italy on the perceived intervals between the descriptive terms, using graphic scaling. A similar study has been carried out in Japan. A major purpose of the study was to examine the uniformity of the perceived intervals in Japanese descriptors, and the validity of this method was verified by subjective evaluation.

View full abstract

This paper describes a new method for the subjective evaluation to image quality using the fuzzy measure. First, some evaluation words are chosen for the structual formula of the model which is used for doing the subjective evaluation to image quality. The structual formula of the model is characterized the membership-function which is in order of the gravity of the evaluation words. Secondly, some observers judge quality of a test image according to the every evaluation word as watching the image on TV monitor, then, these values are exchanged to the fuzzy measure. The numerical values as the fuzzy measure are operated to the membership-function. Then, the total amount of the value of image quality is evaluated as a result of a fuzzy integral with the membership-function and the fuzzy measure.

View full abstract

A new mathematical model of the human vision system has been formulated. This model simulates the human vision spatio-temporal responses quantitatively. The physiological characteristics of the human visual nerve system including the x and Y- ganglion cells have been added to a vision model which was formerly proposed for still images. The new model has been applied to flickering sinusoidal waves, and the model's spatio-temporal frequency characteristics agreed well with experimental results.

View full abstract

It is known that the edge information has an advantage over other depth ques. We perceive a plane even for sparse edge information. Most of traditional methods on edge based disparity detection find the disparity on edge sites, and next, interpolate between them. But, if it fails to correspond edges, the disparity information cannot be corrected later. In this paper we propose an algorithm which detects disparity and, at the same time, interpolate between edge sites. Additionally, we discuss and ability of this algorithm as a vision model.

View full abstract

It is a very important issue to investigate whether right and left eyes gaze the same point in space, when human try to look at a limb of a curved object. So we measured binocular eye movements using two kinds of objects, a cylinder and a prism. For each object, we experimented two cases of whether an object had a texture or not. As a result, we have the following conjectures: In the case of a prism, there seems to be no change in gazing points between the cases of whether an object has a texture or not. But, two eyes seem to gaze much outside for a cylinder with a texture than for the one without a texture, because an occlusion is perceived due to the presence of a texture.

View full abstract

A new type of autostereoscopic 3-D flat-panel display using a LCD-generated barrier-strip method has been developed. This 3-D display employs a couple of LCD panels, the one for displaying computer-generated 3-D images and the other to form a programable parallax barrier-strip. The advantages of the 3-D display include a flexibile display for a desired multi-viewpoint 3-D image, no reduction of resolution in the 2-D image, 2-D / 3-D windowing, and full compatibility with the 2-D image display of personal computers. This autostereoscopic 3-D display has wide applications in science, 3-D computer graphics, education, and amusement as well as 3-D TV use.

View full abstract