Can the Sun‘s Direction be Estimated from an Image Prior to the Computation of Object Shape?

  • Authors:

  • Wojciech Chojnacki;Michael J. Brooks;Daniel Gibbins

  • Affiliations:

  • Department of Computer Science, University of Adelaide, SA 5005, Australia/ and Cooperative Research Centre for Sensor Signal and Information Processing, Technology Park, The Levels, SA 5095, Aus ...;Department of Computer Science, University of Adelaide, SA 5005, Australia/ and Cooperative Research Centre for Sensor Signal and Information Processing, Technology Park, The Levels, SA 5095, Aus ...;Cooperative Research Centre for Sensor Signal and Information Processing, Technology Park, The Levels, SA 5095, Australia/ E-mail: danny@cssip.edu.au

  • Venue:
  • Journal of Mathematical Imaging and Vision
  • Year:
  • 1997

Quantified Score

Hi-index 0.00

Visualization

Abstract

Various computational techniques have been developed that performreasonably well in inferring shape from shading. However, thesetechniques typically require substantial prerequisite information ifthey are to evolve an estimate of surface shape. It is thereforeinteresting to consider how depth might be inferred from shadinginformation without prior knowledge of various scene conditions. Oneapproach has been to undertake a pre-processing step ofestimating the light-source direction, thereby providing input tothe computation of shape from shading. In this paper, we presentevidence that a versatile light-source-direction estimator isunattainable, and propose that, in the absence of domain-specificknowledge, shape and light-source direction should be determined ina coupled manner