Purchase this article with an account.
Takehiro Nagai, Kenji Yokoi, Keiji Uchikawa; Figure segregation achieved by color-distribution differences in colored texture stimuli. Journal of Vision 2003;3(12):79. doi: https://doi.org/10.1167/3.12.79.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
It is well known that luminance or chromatic edges play dominant roles on figure segregation in the visual system. It would be possible, however, that the visual system segregates a figure from its background only by color-distribution differences. We investigated figure segregation properties for two regions having different color distributions in a texture stimulus.
The stimulus (16 deg × 16 deg) was presented on a CRT monitor. It consisted of random shape patches (about 24 min) with gray gaps (3 min) between patches. The stimulus was divided into the test and the background regions. We simulated and interpolated OSA Uniform Color Scales to assign colors of patches. The patch colors in the test and the background regions were selected from two different color distributions, respectively. Each color distribution was of a sphere form with the same radius, but the distribution center of the test region was shifted from that of the background in the OSA color space. We measured the threshold of the distance between the centers in order to segregate correctly the test region from the background region. We used 4 directions of the center shift in the OSA color space. In a trial the subject traced the contour of the test region with a mouse pointer.
The results show that the thresholds are not equal for all directions of the center shift in the OSA color space. Since the OSA color space is perceptually uniform there should be no difference in amount of color-difference in any direction in the space. The present results indicate that the figure segregation is not determined only by color-difference between the test and background color distributions, but is influenced by some other factors in the test and background regions. We discuss about categorical color perception to explain the present results.
This PDF is available to Subscribers Only