Color Recognition in Outdoor Scenes By Non-Parametric Learning

This paper analyzes the variation of the color of objects with respect to existing models ofdaylight and surface reflectance, and shows that the apparent color of objects can be representedby characteristic distributions in RGB space. These distributions can "learned" from trainingsamples using non-parametric approximation techniques. Two approximation techniques -- MultivariateDecision Trees (MDT's) [4] and Neural Networks (with Back-propagation) [36] are usedto approximate decision...