Visually Guided Haptic Search

Systematic Movements in Haptic Search: Spirals, Zigzags, and Parallel Sweeps

Movement strategies were investigated in a one-handed haptic search task in which blindfolded-sighted participants used either one or five fingers to find a landmark on an unstructured tactile map. Search theory predicts that systematic strategies, such as spirals, zigzags, and parallel sweeps, should be more prevalent when the searcher's detection radius is small (one finger) and less common when the detection radius is large (five fingers). As predicted, systematic strategies were more common in one-finger than five-finger search. Participants were able to exploit the larger detection radius during five-finger searches to detect targets with any of their fingers, and in one-finger search used more systematic strategies. For the most part, participants' fingers moved together during five-finger search, expanding and moving quickly when looking for search targets/distractors, and contracting and moving slowly when examining search targets/distractors. There was no evidence of fingers being used as spatial anchors or other independent finger movements in five-finger search. While targets could be found with any fingers, examination was primarily accomplished using the index and middle fingers. Overall, these results indicate that untrained sighted participants will use optimal systematic strategies during haptic search, and this behavior is appropriately modulated by detection radius.

Haptic pop-out of movable stimuli

When, in visual and haptic search, a target is easily found among distractors, this is called a pop-out effect. The target feature is then believed to be salient, and the search is performed in a parallel way. We investigated this effect with movable stimuli in a haptic search task. The task was to find a movable ball among anchored distractors or the other way round. Results show that reaction times were independent of the number of distractors if the movable ball was the target but increased with the number of items if the anchored ball was the target. Analysis of hand movements revealed a parallel search strategy, shorter movement paths, a higher average movement speed, and a narrower direction distribution with the movable target, as compared with a more detailed search for an anchored target. Taken together, these results show that a movable object pops out between anchored objects and this indicates that movability is a salient object feature. Vibratory signals resulting from the movable ball were found to be a reasonable explanation regarding the sensation responsible for the pop-out of movability.

Geodesic Spline Interface for Haptic Curve Rendering

Several haptic devices have been developed in recent years in order to reproduce the sensation of physical contact with virtual objects. Many of these devices are point-based, and some haptic interfaces behave like small surfaces that conform to a virtual shape. None of these allow a full-hand contact with the shape, and they are, in general, too small to render big surfaces. The simulation of tasks, such as the exploration of aesthetic surfaces made by industrial designers in order to check the quality of prototypes, require full-hand contact with the shape on a one-to-one scaled representation of the object. These explorations follow trajectories that can be approximated with planar or geodesic curves. In this paper, we describe the design and implementation of a linear haptic device that is able to render these trajectories. The device is part of a multimodal system including stereoscopic visualization that allows visual representation of the entire surface. Industrial designers use the system for checking the quality of shapes while exploiting their manual and visual skills. The system has been tested by industrial designers and the results are reported in this paper.