Academic journal article Human Factors

Visual Context and the Control of Movements through Video Display

Academic journal article Human Factors

Visual Context and the Control of Movements through Video Display

Article excerpt

The aim of this experiment was to test the influence of target context on adaptation to scale perturbations introduced by a video display. Participants performed pointing movements without direct vision of their moving hand, although they could see their movements on a video display. Their perceived movements could be reduced, enlarged, or displayed at their actual size. Three target contexts were compared: dark surround, illuminated frame, and familiar object. Movements were executed with or without vision of hand displacement. Results showed that target context enhanced an allocentric coding of the movement, which improved movement execution. However, the effect of target context changed whether or not the displacement of the hand was available. Overall, the results suggest that target context allowed the extraction of dynamic information about movements, which is used to program and control movements. This suggests that target context could be used efficiently to improve spatial accuracy and speed in teleo peration learning. Potential applications include the reduction of difficulties encountered during teleoperation learning through the introduction of visual context.

INTRODUCTION

Optimal visuomanual coordination requires that visual and proprioceptive information about movement be complete and concordant (Jeannerod, 1988). However, such information is often degraded because of the introduction of indirect modes of control. For example, teleoperation emerged from work in hostile environments, such as space and oceans. Teleoperation involves the remote control of a task by means of a teleoperator (robot or articulated arm) and a camera that transmits images of the position and displacement of the teleoperator. Although this process is mostly used in hostile environments, it can also be exploited in restricted areas, such as telesurgery. This minimally invasive surgery involves making small incisions through which a camera and tools are introduced. A robot can also be used to perform operations remotely. Telesurgery and robotic surgery are now in wide use. However, remote manipulations are likely to raise problems with neuromuscular control of the operator.

Teleoperation not only introduces a discrepancy between vision and proprioception, it also leads to specific visual perturbations. The viewpoint of the camera becomes dissociated from the viewpoint of the operator. The camera/operator discrepancy often leads to direction and scaling problems. Finally, the information commonly used to perceive the distance of objects is disturbed (see Reinhardt-Rutland, 1996, for a review). Monocular information (e.g., accommodation) and binocular information (e.g., convergence and binocular disparity) are useless for depth perception on a flat screen. Moreover, motion information is often ineffective during performance of a remote task, given that the visual field is generally reduced.

The image is sometimes enlarged in a remote task, yielding a reduction of the visual field and an erroneous perception of movement velocity. For example, a paramecium seen through a microscope seems to move quickly, relative to reality (Lumsden, 1980). Consequently, only pictorial cues such as interposition, texture, and familiar size of an object provide a veridical image that can help to guide movements via video display (Hagen, Click, & Morse, 1978). In addition, zooming in and Out is known to greatly disturb the evaluation of an object's distance, and only a relative depth between two objects can be estimated adequately (Reinhardt-Rutland, 1996). Moreover, pictorial information is adequate only when objects are well defined, which is problematic in telesurgery, given the nature of body tissue. Indeed, in spite of precise anatomical representation of the human body, surgeons are working with blurred contours belonging to diverse building strata. It can be seen, therefore, that a precise evaluation of the v isual scene is very difficult in remote operations. …

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.