Perception of illusory contour and spatio-temporal integration in the visual system

Abstract

Spatial and temporal conditions to perceive an illusory contour figure were examined. Six figural elements inducing an illusory figure were successively presented. Total time to present all inducers and the ratio of duration to 1S1 between inducers were systematically varied. Five subjects were required to rate the clarity of illusory contour, brightness difference and apparent depth difference on an 11 point scale. The effects of total time and the ratio of duration to ISI were significant. The results of causal analysis indicated that total time determined contour, brightness and depth. In two subjects durationfISI had an influence on brightness. Contour and depth were closely related in all subjects. Spatio-temporal integration in the visual system which is higher than in the VIS was discussed.

Spatiotemporal integration and contour interpolation revealed by a dot localization task with serial presentation paradigm

“…Taking into consideration the difference in the temporal characteristics of the contour system and the surface‐spreading system (Ringach & Shapley, 1996), the surface‐spreading system might operate in the 240‐ms condition. The notion of the difference of the temporal characteristics of the contour system and the surface‐spreading system is consistent with the estimates of spatiotemporal integration, approximately 300 ms, using brightness change as a measure (Takemoto & Ejima, 1997; Unuma & Tozawa, 1994). The unit representation of objects (Kellman et al., 1998, 2001) might be related to such a surface representation.…”

“…There is evidence that spatiotemporal integration is temporally limited to relatively brief time intervals (Shipley & Kellman, 1994; Takemoto & Ejima, 1997; Unuma & Tozawa, 1994). For instance, Shipley and Kellman (1994) examined boundary formation and shape identification using spatiotemporal transformations of spatially separated elements, and found that spatiotemporal integration appears to be less efficient when temporal ranges exceed 165 ms. Unuma and Tozawa (1994) used a serial presentation paradigm in which each inducing part of an illusory contour figure was presented individually in a particular serial order, and analyzed causal relations between perceived brightness‐change, contour‐clarity, and depth‐change. They found that an illusory contour figure could be perceived from fragmented edges within a total temporal range that was greater than 260 ms, as given by the stimulus‐onset asynchrony (SOA) between two successive inducing elements.…”

“…They found that an illusory contour figure could be perceived from fragmented edges within a total temporal range that was greater than 260 ms, as given by the stimulus‐onset asynchrony (SOA) between two successive inducing elements. In light of this, Unuma and Tozawa (1994) suggested that illusory contours can be perceived from fragmented elements over relatively long temporal ranges, and that the perception of brightness change appears to be independent of the perception of illusory contour. The fact that estimates of temporal limits appear to vary across experimental reports suggests that the time required for spatiotemporal integration may be influenced by several factors, such as task demands, stimulus attributes, observers' knowledge, and the particular behavioral measures employed.…”

“…Assuming that the temporal characteristics of spatiotemporal integration of visible edges, fragmented both in space and time, are closely related to the microgenetic process (Sekuler & Palmer, 1992) of spatial interpolation with static displays, contour interpolation seems to develop within 200 ms (Keane et al., 2007; Ringach & Shapley, 1996; Shipley & Kellman, 1994). In contrast, surface processes have been proposed to require further processing time (Ringach & Shapley, 1996; Sekuler & Palmer, 1992; Takemoto & Ejima, 1997; Unuma & Tozawa, 1994). Therefore, the present research aims to clarify the fidelity of interpolated contours with spatiotemporal integration that entails two different temporal phases.…”

“…This is the short‐range condition. The total time in Experiment 2 ranged from 240 to 1920 ms (Takemoto & Ejima, 1997; Unuma & Tozawa, 1994); this is referred to as the long‐range condition.…”