著者
黒田 正巳
出版者
一般社団法人日本建築学会
雑誌
日本建築学会論文報告集 (ISSN:03871185)
巻号頁・発行日
no.55, pp.84-93, 1957-03-25

Correct retinal image cannot be drawn, but several assumtions maxe it possible. They are-fixed mean sight line, flat projection plane, no peripheral vague image, no binocular double image, and no peripheral curved image. Under these assumptions I deviced one of 3-vanishing points perspective drawing methods, and adopted it as an "approximate pure retinal image." Retinal image vary with fixation. Question is whether perceptive image does so. Subjects are let to observe a square figure (2.5m×2.5m) on a vertical plane, fixating their sight lines on several points (fixation points), and they are instructed to draw their perceptive images. Those drawings were compared with those retinal images. Main conclusions are: 1. In spite of pretty so much individual differences, mean perceptive images are almost constant wherever existing the fixation points. 2. Perceptive images converge toward horizontal direction, a little converge toward vertical direction. 3. The smaller deflection angle of observation point and the smaller degree of constancy.

言及状況

Twitter (2 users, 2 posts, 1 favorites)

恒常視はお絵かきには役に立つ気がする:CiNii 論文 -  建築における形の恒常視に関する実験的研究 : 第二、網膜像の作図及び凝視点がある場合の知覚像 [ http://ci.nii.ac.jp/naid/110004853346/ ]

収集済み URL リスト