Constraints on long range interactions mediating contour detection

Abstract

Contour detection may be mediated by lateral interactions between neighboring cortical neurons whose receptive fields have collinear axes of preferred orientation. This hypothesis was tested in psychophysical experiments and computer simulations using a contour detection task in which observers searched for groups of Gabor patches that followed spatially extended contour paths embedded in noise consisting of several hundred Gabor patches with random positions and orientations. The orientation-selective units in the simulated neural network were linked by facilitatory interconnections whose strength depended on the geometry (distance, curvature, change in curvature) of smooth curves connecting the orientation axes of units in a pairwise fashion. Psychophysical detection performance was much higher for contour signal groups that followed closed rather than open-ended paths. However, just two sudden changes in orientation of neighboring Gabor patch elements in closed-path contours reduced detection performance to the same levels obtained with open-ended contours. These psychophysical data agreed with the results of the neural network simulations. Furthermore, the simulations also accounted for previous findings that removal of a single Gabor patch element from a closed-path contour group significantly degraded detection performance. We conclude that closure alone is not sufficient to enhance the visibility of a contour. However, if a closed contour meets certain geometric constraints, then lateral interactions based on these constraints can generate facilitation that reverberates around the closed path, thereby enhancing the contour's visibility.

How long range is contour integration in human color vision?

“…tal results (Kovacs & Julesz, 1993;Pettet et al, 1998). In particular, Kovacs and Julesz (1993) reported that the critical separation for closed contours is about two times larger than for open contours (for elements of relatively high spatial frequency, 8.3 cpd).…”

“…In particular, Kovacs and Julesz (1993) reported that the critical separation for closed contours is about two times larger than for open contours (for elements of relatively high spatial frequency, 8.3 cpd). However, in these previous experiments the relative density between contour and background elements was changed across conditions, with either the spacing between contour elements varied and the spacing between background elements fixed (Kovacs & Julesz, 1993), or the spacing between contour elements fixed and the number of background elements varied (Pettet et al, 1998). The combination of the local density cue present in their stimuli and contour closure could thus explain the facilitation reported for the detection of closed contours.…”

Perceptually Motivated Image Features Using Contours

“…Two categories were motivated by the information commonly used by existing texture features: 2nd-order statistics, and short-range higher order statistics (HOS) (typically obtained from image patches). The use of the third category, contour data, was motivated by the fact that the human visual system is extremely adept at exploiting these visual cues [25], [44], [46][47], [53] and that they utilise long-range HOS. We conducted an experiment with human observers that showed that for the Pertex database [29], contours are the most useful category of data for human texture discrimination.…”

“…These images are outlined using the above colour scheme to indicate which property is most important to their recognition: contour maps (red), 2nd-order statistics contained in phase-randomised images (blue) or the short-range interactions contained in randomised blocked images (cyan). [46][47]. In addition, it has been shown that objects can be identified using discontinuous fragmented contour segments [44].…”

“…It is well-known, however, that these characteristics have an important role in human visual perception [15], [25], [42], [44], [53], [55]. For example, human observers often cannot recognise an aperiodic image when its phase spectrum is scrambled and its power spectrum is kept intact [42], but they are able to exploit the long-range visual interactions evident in contour information [25], [44], [46][47], [53]. Design of perceptually inspired computer algorithms has been studied in the community [2] but, to our knowledge, no research has been reported which utilises contour data 2 for texture analysis.…”