The shape of road markings for visibility computations

Road markings visibility is usually assessed by means of the visibility level (VL). Several models have been developed to calculate the visibility distance on the road (Detect/PCDetect, Carve/Tarvip, Cost 331 and CIE 73). These models all approximate the visibility of a marking by estimating the diameter of an 'equivalent' disc. In this process, speculative hypotheses are made to calculate this diameter, labelled the equivalent area hypothesis, the equivalent perimeter hypothesis or the equivalent width hypothesis. To compare these hypotheses, a psychophysical experiment was conducted with eight participants and 120 stimuli. It was found that the merit of each hypothesis changed with the diameter of the equivalent disc. Above 0 center dot 1 degrees, all the hypotheses gave good results, but the equivalent area hypothesis was found to be the best. Below 0 center dot 1 degrees, the equivalent width hypothesis was more accurate, but further investigations are still needed.