Purchase this article with an account.
Harold T. Nefs, Jan J. Koenderink, Astrid M. L. Kappers; Shape-from-shading for matte and glossy objects. Journal of Vision 2004;4(8):473. doi: 10.1167/4.8.473.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
It has been claimed many times before, that the addition of specular highlights to otherwise matte objects alters the 3D-perception of those objects. However, no extensive scientific data has ever been published to support this claim. We were therefore interested to find out whether the presence of specular highlights on glossy objects would make a difference in surface relief if compared with matte, i.e. Lambertian objects. Shape-from-shading was evaluated for matte and glossy objects with two different paradigms. We used images of six different curved globally convex objects, which we displayed on a computer screen. The depicted objects were either matte or glossy and were illuminated from one of two different directions. In Experiment 1, four observers were asked to set a number of local surface attitude probes such that they looked tangent to the objects' surfaces. In Experiment 2, three other observers were instructed to make traces of the contour of the depicted objects in the transversal and sagittal planes. There are important differences between the two paradigms. Whereas the settings in the Experiment 1 reflect depth gradients, Experiment 2 targets depth directly. Furthermore, the paradigm in Experiment 1 is often considered to be a task involving local shape estimates whereas the paradigm in Experiment 2 is a task that relies more on global shape estimates. Surprisingly, in both experiments little difference was found between the results for matte and glossy objects. However, differences in surface relief were obviously present for different illumination directions. From the results of these two experiments we must conclude that there is no sign that glossiness influences shape perception although the appearances of matte and glossy objects are quite different. We also conclude that observers must thus be able to disregard specular highlights and do not misinterpret them as areas where the illumination is perpendicular to the surface.
This PDF is available to Subscribers Only