Purchase this article with an account.
Hongfan Shen, Damien Mannion, Seong-Whan Lee, Daniel Kersten; Accuracy in Localising the Centre of a Circle. Journal of Vision 2014;14(10):773. doi: 10.1167/14.10.773.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In everyday life, humans can point to, or aim at, the centre of an object with apparent ease. However, accurately locating an object's centre is computationally challenging due to the inherent complexity of the object's representation in the visual image and the spatial uncertainty of the object's boundaries. Here, we probed the accuracy of human observers in estimating the centre of a simple circle. Participants (n=10) were presented with a circular contour defined by a set of dots, and were asked to translate the dots such that the centre of the circle was aligned with a visible target point. To probe observer prformance, we manipulated the number of dots defining the contour (8, 32, or 128), the size of the circle (3, 6, or 18 deg visual angle), and the level of radial position noise applied to the dots (drawn from a Gaussian with a standard deviation of 1%, 3%, 10%, or 30% of the circle radius). Observer estimation error was well captured by a function that quantified performance in terms of the radial position noise in the stimulus, the observer's internal noise, and the number of dots utilised by the observer. We find that the estimates of internal noise increased approximately linearly with increasing circle size, while being roughly equivalent for different numbers of dots. This appears consistent with an increased positional uncertainty in the periphery of the visual field. Conversely, the estimates of the number of utilised samples increased with number of dots but was roughly constant across circle size. Overall, this study represents a quantification of human performance on a simple centroid localisation task that will allow for future comparisons with ideal and sub-ideal observers to understand the computational underpinnings of this important visual capacity.
Meeting abstract presented at VSS 2014
This PDF is available to Subscribers Only