Abstract
When visual information enters the brain, segregating scenes into figures and background is one of the first processes established by the visual system. Previous research has shown that early visual areas are engaged in low-level feature detection, and probably also in sub-processes related to texture segregation, such as the detection of orientation discontinuities. We conjecture that in addition, V1 and other early areas are involved in the perception of the surface layout, i.e. the final stage of figure-ground segregation. To test this, we manipulated the perceptibility of a texture-defined figure by interposing a ’frame’ between figure and ground textures. For example, the orientation of the figure texture was 45o, the background texture was 135o, and the intervening frame texture was 90o. Subjects had to distinguish between figure-ground textures surrounded by a frame and isolated frames on a homogenously textured background. Subjects could easily discriminate between the two when the frame thickness was relatively small, yet were strongly impaired when the frame thickness was relatively large. Apparently, in the latter condition the proper perception of the different texture surfaces is lost. We measured BOLD activity during the performance of this task. Preliminary results showed a direct parametric relation between frame thickness on the one hand and the BOLD contrast between figures surrounded by a frame and isolated frames on the other hand, in areas V1, V2, and V3. This indicates that activity in these early visual areas reflects the perceived surface layout of texture displays. This supports a role of these areas in perception, in addition to their established role in lower level feature detection.