Foreshortening produces errors in the perception of angles pictured as on the ground

Abstract

Observers viewed pictures of a simulated ground plane and judged the orientation of lines pictured as lying on the ground. We presented three lines at a time and manipulated three factors: (1) the declination of the lines below the horizon (depicting distance to the target angles), (2) their length, and (3) whether or not they converged to a point on the horizon. Only the first factor had a substantial effect on these errors. We conclude that perspective foreshortening in pictures produces errors in perceived 3-D orientation. Our explanation is based on the different rates of change of elevation and azimuth with distance.

Inverse Sideways Perspective on a Road: A Drawing by a Blind Adult

“…The space between ties also subtends an angle, in elevation. This elevation angle also diminishes with distance but at a quadratic rate Wnuczko et al, 2016). The linear/quadratic relation between azimuth and elevation angles is not understood by the vast majority drawing freehand.…”

Flying-Gallop and Foreshortening: Winner Circle

“…However, if the terrain appears compressed with distance, the erstwhile diagonals would be seen meeting the scene's z-line at 60 or 80 and more. Wnuczko, Singh, and Kennedy (2016) offered viewers pictures of a textured plain (Figure 7). They judged the orientation to the z-line of oblique lines on the ground, shown by lines in the picture converging to a common point on the horizon.…”

“…Wnuczko, Singh, and Kennedy (2016) offered viewers pictures of a textured plain (Figure 7). They judged the orientation to the z -line of oblique lines on the ground, shown by lines in the picture converging to a common point on the horizon.…”

“…Acute angles made with the z -line appear to increase with elevation. Reprinted with permission from Wnuczko et al. (2016).…”

Pot/Lid Illusion

“…Visual “overconstancy” of this kind has been reported, but rarely, and only for extremes of perspective foreshortening ( Juricevic & Kennedy, 2006 ). Rather, the standard argument about optic-slant is that we judge modestly foreshortened sizes correctly and when optic slant produces foreshortening of 80% and more, as optic slant approaches its upper limit, stretches of ground look shorter than true ( Wnuczko, Singh, & Kennedy, 2016 ). Nevertheless, we mention the logical alternative that moderate foreshortening such as the 40% in Figure 1 , given 60° optic slant, may have elongated apparent z-dimensions compared with vertical dimensions.…”