August 2009
Volume 9, Issue 8
Vision Sciences Society Annual Meeting Abstract  |   August 2009
Memory colours of polychromatic objects
Author Affiliations
  • Milena Vurro
    Institute of Neuroscience, Newcastle University(UK)
  • Yazhu Ling
    Institute of Neuroscience, Newcastle University(UK)
  • Anya Hurlbert
    Institute of Neuroscience, Newcastle University(UK)
Journal of Vision August 2009, Vol.9, 333. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Milena Vurro, Yazhu Ling, Anya Hurlbert; Memory colours of polychromatic objects. Journal of Vision 2009;9(8):333.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Memory colours of familiar objects may assist in colour constancy and object recognition, as suggested by Hering (1874). Natural objects are polychromatic (Hurlbert et. al., VSS 2007); therefore memory colours may be intrinsically variable, influencing their role in constancy. Here we study the effect of shape and natural chromatic texture on the accuracy and variability of memory colours for natural objects.

We use a novel set-up that allows us to manipulate the non-uniform surface reflectance properties of solid objects under controlled illumination. Using a tristimulus-calibrated camera, we obtain images of real fruits and vegetables under characterized illuminations, eliminating shading and highlights. We morph the flattened tristimulus image to fit the visible surface of matching solid objects (painted white), maintaining the geometry, surface chromaticity and luminance values of the real objects.

Under controlled adaptation conditions, observers freely viewed individual objects. Their task was to adjust the ‘colour’ of the object until it appeared ‘natural’ for the depicted or stated object, using a joystick. We varied object shape (2D filled natural outline, 3D natural shape, or generic rhomboid shape (2D or 3D)), chromatic texture (natural polychromatic texture or single uniform chromaticity), and illumination. In adjusting the ‘colour’, observers freely rotated the mean angle of the chromaticity distribution (polychromatic textures) or the single chromaticity (chosen as the mean or maximally saturated of the natural distribution).

The mean accuracy memory colour tends to be higher, and the error range is significantly smaller, for surface colours attached to their natural 3D shapes, in comparison with other shape conditions, for both polychromatic and uniform conditions. Polychromaticity (vs. uniform chromaticity) improves accuracy for natural 3D shapes, but the effect depends on object type and illumination. We conclude that memory colour is linked to shape in object representation, and that polychromaticity does not influence its variability.

Vurro, M. Ling, Y. Hurlbert, A. (2009). Memory colours of polychromatic objects [Abstract]. Journal of Vision, 9(8):333, 333a,, doi:10.1167/9.8.333. [CrossRef]
 Supported by EPSRC EP/D068738.

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.