Abstract
The paper reports experimental work to check the validity and significance of Keck's proposed criteria for the assessment of road lighting installations. The experimental site was in Rouen, lit on one side by five semi-cutoff lanterns whose height and spacing can be varied. 21 observers estimated the visibility level of a target (20cm × 20 cm) for various positions and spacing / height ratios of the luminaires. The averages of the responses on a scale 1 to 4 are compared with visibility levels calculated with either experimental or theoretical luminance values. In both cases the optimum correlation (r = 0.94) was obtained when taking as the contrast value the ratio between the target luminance and the average of the base light intensities at locations to the left and right of the target. Moreover, there is no difference between experimental and theoretical results. Conclusionsare drawn as to the significance and drawbacks of target visibility levels for the assessment of road lighting.