3D scenes are often presented as pictures on desktop monitors. Much of the work on size perception in pictures has used measures involving visual matching. How observers perceive pictorial displays with respect to body-based judgments about action capabilities has not been examined, and may be relevant for applications intended to display act-on-able objects or environments such as in architectural design. Affordance judgments have been used in real and immersive environments, often focused on larger scale actions such as judgments of passage. Here, we use judgments of the ability to grasp a cube to test size perception on a computer desktop display. In the real environment, participants viewed the cubes on a table. In the graphics display, participants viewed a rendering of the same cubes and table on a desktop display. In both conditions, cubes were placed 50 cm and 70 cm from the participant. The cubes were viewed binocularly from a viewpoint location that matched the rendering location. Results were analyzed as a ratio of judged over actual ability. Results revealed a main effect of viewing environment. Participants in the desktop display judged that they could pick up larger cubes than when in the real environment. There was also an interaction of viewing environment and location. The effect of distance was greater in the desktop condition than the real environment. The desktop results are consistent with the size-distance invariance hypothesis, modified by the presumption that distance perception in pictures is affected by both the pictorial cues for distance and the distance of the screen. The real world results show that in the absence of a screen, the judgments are conservative and are minimally affected by distance to the objects. Work is underway to confirm these results with a reaching through measure.
Meeting abstract presented at VSS 2012