Browsing by Subject "color vision"

An object’s distinct color is commonly formed from a chromatic distribution, with varying hue information. Human observers can estimate the mean hue of a hue ensemble, but the details of this ability are still relatively unknown. Ensemble perception of hue was studied by systematically varying the external noise and the amount of information available. In a set of four experiments, the spatial sampling characteristics, spontaneous estimation strategy and task dependence, were examined. Presented briefly against a gray background, the stimuli consisted of a number of square elements, each with a uniform hue. The hues were drawn either from a von Mises distribution or a highly skewed normal distribution with one of three levels of external noise, located on a hue circle in CIELAB color space. In Experiments 1-3 observers performed a two-interval forced choice task and in Experiment 4 a modified match-to-sample task. Discrimination thresholds were estimated by fitting psychometric functions to the data. The number of elements utilized in averaging was estimated through equivalent noise modeling. Discrimination thresholds increased with increasing external noise, but decreased as the number of elements increased, the improvement being greater with higher noise. Modeling the number of samples used by the observer as a fixed power of the samples available gave an excellent fit to the data. For the 64-element stimulus, estimated effective number of samples ranged from 16 to 41. Control experiments confirmed that performance improved with the number of elements, not stimulus area. Even with a highly skewed hue distribution, most observers opted for a simple averaging strategy in estimation. With a less straightforward comparison task, results appeared much less consistent. Observers sample and average stimulus elements to estimate mean hue, similarly to ensemble perception in other domains. The observed sampling clearly surpasses earlier estimates pointing towards global sampling, and performance is most affected by surface, not edge information. Simple averaging remained as the observers’ spontaneous strategy with a non-normal hue distribution. The results were also dependent on task details, emphasizing caution when comparing different kinds of experiments.