Road sign recognition: a study of vision-based decision making for road environment recognition

This study is based on an application of vertical road sign recognition by vision. Three types of danger warning signs are recognized. Octagonal stop signs and triangular danger warning signs are distinguished from round forbidding signs on the basis of their outside shape. This recognition is made in “quasi-real-time” in a running vehicle. In addition to the vision algorithm strategy developed, three decision-making softwares are compared: structured programming, expert system approach, and a neural network.