抄録
Role of vergence signal in visual size perception has been investigated using a computer generated 3D pattern projection system. Two horizontally separated red dots(10cm diameter)were projected alternately with 48Hz on a large front-parallel transparent screen(2.5m × 2.5m)from behind, which was 50cm from a subject. Wearing a binocular liquid crystal shutter, a subject tried to reproduce the previous dot size with a joy stick, after the size, inter-dot separation, or both were changed suddenly by the computer. It has been proved that all of the subjects investigated(7 male, 1 female)perceived shrinking of the dot as vergence angle was increased. Thus, the subjects tried to enlarge the dot to reproduce the initial objective size. The extent of enlargement depended linearly on the dioptric distance of vergence. However, the size of the enlarged dot was much smaller than the value theoretically predicted, and also showed big individual difference(9〜68% of theoretical prediction)among subjects. This implies a possibility that vergence signal alone cannot play significant role on many subjects in determining object distance with exceptions of restricted number of subjects. On the other hand, a possibility remains that vergence signal is correctly estimated, while expression of size via visual system is incorrect. To elucdate the possibility, the subjects were instructed to image suitable preshaping of the hand invisible to grasp the object just before adjusting the pattern size. Two subjects clearly improved the performance of size reproduction (from 68% to 105%, and 10% to 48%), while other 3 subjects showed no improvement. The finding implies that some subjects estimate vergence rather correctly, and can estimate the objective pattern size with the aid of imagination of grasping the object. Other 3 subjects seem to have poor estimation ability of vergence. In order to test the role of vergence in determining object distance, subjects were instructed to reproduce initial vergence distance by adjusting the inter-dot separation after it together with dot size were changed. Those who demonstrated good reproducibility of dot size again proved to be able to reproduce vergence distance correctly, while others who had poor reproducibility of size showed poor ability of reproducing the initial vergence distance.
