The emergence of road-following as a machine vision application has spawned several methods of utilizing color to enable autonomous vehicles drive without specific parametric models. Crisman's SCARF road-following algorithm  approximates an ``average'' road color from samples, and models the variation of the color of the road under daylight as a Gaussian distribution about an ``average'' road color, and classifies pixels based on minimum-distance likelihood. Pomerleau's ALVINN road-follower  uses color images of road scenes along with user-induced steering signals to train a neural network to follow road/lane markers. Although the ALVINN algorithm made no attempt to explicitly recognize lanes or roads, it showed for the first time, that a complex visual domain with unmodeled variation can be approached as a non-parametric learning problem. These techniques were successfully applied to road-following, but are not meant for general color recognition.