Computational models of visual neurons specialised in the detection of periodic and aperiodic oriented visual stimuli: bar and grating cells

Abstract

Abstract. Computational models of periodic-and aperiodicpattern selective cells, also called grating and bar cells, respectively, are proposed. Grating cells are found in areas V1 and V2 of the visual cortex of monkeys and respond strongly to bar gratings of a given orientation and periodicity but very weakly or not at all to single bars. This nonlinear behaviour, which is quite different from the spatial frequency filtering behaviour exhibited by the other types of orientation-selective neurons such as the simple cells, is incorporated in the proposed computational model by using an AND-type non-linearity to combine the responses of simple cells with symmetric receptive field profiles and opposite polarities.

Suppression of contour perception by band-limited noise and its relation to nonclassical receptive field inhibition

“…The suppression effect they observed was strongest when the surround grating and the optimal bar stimulus over the CRF had the same orientation and decreased when the angle between the two was increased. This dependence on the orientation difference is well reproduced by a computational model of this type of cell proposed by Petkov and Kruizinga (1997).…”

“…(Daugman 1985). We use a modified parameterization according to Petkov and Kruizinga (1997) that takes into account restrictions found in experimental data. A receptive field function g k;r;h;u ðx; yÞ; ðx; yÞ 2 X & R 2 , centered in the origin specifies the response to an impulse at point ðx; yÞ and is defined as follows: g k;r;h;u ðx; yÞ ¼ e Àx x 2 þðcỹ yÞ 2 2r 2 cosð2px k…”

“…The model used in (Petkov and Kruizinga 1997) also involves half-wave rectification and contrast normalization, but we do not need these aspects of the function of simple cells in the context of this paper. In image processing and computer vision, the filter defined by Eqs.…”

“…On the horizontal axis the noise frequency ranges from 0.38 to 3.05 cycles per degree (cpd) Elsewhere, the perceptual effect illustrated by Fig. 2c was related to the function of a specific type of orientation-selective cell in the primary visual cortex for which the response to an optimal stimulus over the cell's classical (excitatory) receptive field (CRF) is influenced by other stimuli outside that area (Petkov and Kruizinga 1997). In contrast to other orientation-selective neurons, such as simple (Hubel and Wiesel 1968), complex (Hubel and Wiesel 1968), and grating (von der Heydt et al 1992) cells, the concerned type of neuron responds strongly to single bars (of appropriate size and orientation) over its CRF and shows reduced response when a grating is added to the surround.…”

Interaction of ON and OFF pathways for visual contrast measurement

“…This has been interpreted as evidence for non-linear sub-®eld integration, since the responses could not be reconciled with a linear ®lter model responsive to gratings of the given spatial frequency. The responses were critically dependent on the precise stimulus periodicity, suggesting that the cells integrate spatially aligned arrangements of alternating ON and OFF LGN input in a non-linear way [see von der Heydt (1987) for an overview and discussion of previous ®ndings in that direction, and Petkov and Kruizinga (1997) for an approach to model the key behavior].…”