Academic journal article Attention, Perception and Psychophysics

Visual Space Perception at Different Levels of Depth Description

Academic journal article Attention, Perception and Psychophysics

Visual Space Perception at Different Levels of Depth Description

Article excerpt

Published online: 6 May 2015

© The Psychonomic Society, Inc. 2015

Abstract The main purpose of this study was to determine the effect of the depth description levels required in experimental tasks on visual space perception. Six observers assessed the locations of 11 posts by determining a distance ranking order, comparing the distances between posts with a reference unit, and estimating the absolute distances between the posts. The experiments were performed in an open outdoor field under normal daylight conditions with posts at distances ranging from 2 to 12 m. To directly assess and compare the observers' perceptual performance in all three phases of the experiment, the raw data were transformed to common measurement levels. A pairwise comparison analysis provided nonmetric information regarding the observers' relative distance judgments, and a multidimensional-scaling procedure provided metric information regarding the relationship between a perceived spatial layout and the layout of the actual scene. The common finding in all of the analyses was that the precision and consistency of the observers' ordinal distance judgments were greater than those of their ratio distance judgments, which were, in turn, greater than the precision and consistency of their absolute-magnitude distance judgments. Our findings raise questions regarding the ecological validity of standard experimental tasks.

Keywords Depth perception · Visual space · Level of description · Distance order · Depth scales · Exocentric distance

How often does one need to determine the depth relationships between surrounding objects with high accuracy and discriminate fine details? How often does one need to determine the metric properties of a 3-D scene? Somewhat surprisingly, such requirements are only seldom met. In many real-life situations, the observer only makes nominal or ordinal judgments of a scene's spatial layout. When deciding how to interact in an environment on the basis of available sensory data, the observer often makes a discrete choice, such as when attempting to answer many types of questions: Should I begin running to catch the bus that has just arrived at the bus stop? Can I cross the road right now without the risk of being hit by an approaching car? Which of the available paths will get me to my destination most quickly? Could I manage to hit that fast-moving tennis ball? Will I be able to reach the peak of that mountain on the horizon before sunset? None of these spatial tasks is overly complex, and highly detailed reconstruction of depth relationships is not needed in these cases. Thus, from an evolutionary perspective, it seems plausible to assume that information about spatial relationships is determined in a quantity and quality sufficient for solving the task and that the observer 's visual system uses simple, direct strategies (e.g., Gibson & Bergman, 1954; Glennerster, Rogers, & Bradshaw, 1996;Lee,1980). For instance, Gibson (1950)argued that much of our perceptual awareness of the environment is based on simple order relations that can be described in terms of Bgreater than^ or Bless than.^ According to both Cutting and colleagues and Koenderink and colleagues, perceived depth is normally confined to a number of depth planes; only when the observer needs to know more precise information about how an object or its parts are positioned in space (such as when recognizing a face) does perceived depth become more articulated (Cutting, 1998, 2003; Cutting & Vishton, 1995; Koenderink, van Doorn, & Wagemans, 2011; van Doorn, Koenderink, & Wagemans, 2011).

Contrary to real-life experience, most research on visual space perception has required an observer to provide a metrical description in terms of Euclidean distances and angles (for exceptions, see Aznar-Casanova, Matsushima, Da Silva, & Ribeiro-Filho, 2008; Norman & Raines, 2002; Norman & Todd, 1998; Norman, Todd, Perotti, & Tittle, 1996;Todd& Reichel, 1989; Toye, 1986; van Doorn et al. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.