Academic journal article Perception and Psychophysics

Roughness Perception in Haptic Virtual Reality for Sighted and Blind People

Academic journal article Perception and Psychophysics

Roughness Perception in Haptic Virtual Reality for Sighted and Blind People

Article excerpt

Psychophysical functions for perceived roughness, relating ln (magnitude estimate of roughness) to ln (groove width), were obtained for blind and sighted participants in virtual reality using the PHANToM force feedback device. The stimuli were sinusoidal surfaces with groove widths between 0.675 mm and 20.700 mm. Group functions showed a similar nonlinearity to those obtained in physical reality using rigid probes (Klatzky, Lederman, Hamilton, Grindley, & Swendsen, 2003; Lederman, Klatzky, Hamilton, & Ramsay, 1999). Individual functions gave a different picture. Of 23 total participants, there were 13 with wholly descending linear psychometric functions, 7 with quadratic functions similar to the group function, and 3 with anomalous functions. Individual power law exponents showed no significant effects of visual status. All analyses gave a power law exponent close to -0.80. The implications for theories of roughness, methodologies of data analysis, and the design of haptic virtual reality interfaces are considered.

Roughness is one of the primary perceptual properties of objects. For example, Linus's attachment to his soft security blanket may evoke empathy even in the most scientific of psychophysicists. Indeed, the physical properties that determine roughness have long been of psychophysical interest. Advances in technology have added to our knowledge in several ways. There is now a wider range of well-specified physical stimuli. Research is no longer limited to running fingers over sandpaper of different grain sizes. Nevertheless, the spacing and size of the texture elements on a surface remain the primary, if not the only, physical variables of interest. Consequently, determining the psychophysical function that relates perceived roughness to interelement spacing remains a major focus of research. Furthermore, it is now possible to produce virtual stimuli with properties that parallel those of the physical world. These developments have the potential to provide tactile interfaces for everyone, and they offer particularly exciting opportunities to improve interfaces for blind and partially sighted people. The latter is a particularly important issue, since interfaces that replace visual information with auditory and haptic information are greatly needed by blind and partially sighted people (Petrie, O'Neill, & Cohvell, 2002). Although these virtual reality devices have been developed primarily for such practical applications, they also provide potential for theoretical psychophysical investigations.

The main aim of this study was to take advantage of these advances to obtain psychophysical roughness functions in virtual reality using a force feedback device. These psychophysical functions have implications for whether a single or a dual process underlies roughness perception; they also enable investigation of differences in virtual roughness perception between blind and sighted people, with implications for the extent to which visual experience influences touch perception. Psychophysical functions can also be used to address a range of applied issues. Differences in psychological function that result from type of probe or participants' visual status have implications for the design of tactile interfaces to many applications. A further ami of this study was to address the methodological problems that arise from differing traditions for studying roughness perception. We believe this study is unique in providing both individual analyses, in the tradition of classical psychophysics (Stevens & Galanter, 1957), and group analyses using ANOVA techniques (Lederman & Taylor, 1972). Because the psychophysical functions may be nonlinear, the two methods can (and do) lead to different conclusions.

Pioneering work in the psychophysical tradition used magnitude estimation to obtain the Stevens's power law exponent for roughness for individual participants. The first studies used sandpaper stimuli (Brown, 1960; Ekman, Hosman, & Lindstrom, 1965; Marks & Cain, 1972; Stevens & Galanter, 1957; Stevens & Harris, 1962; Stone, 1967) and found that roughness increased as granule size increased (increased granule size corresponds to increased interelement spacing, where granules are the elements). …

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.