Abstract
Road visibility information is important for driving safety, and the need for such information in winter is particularly great. Road administrators provide drivers with road images gathered using closed-circuit television (CCTV) cameras installed along national highways. However, when road condition information is given as an image, the provision media are limited. Japan's Hokkaido Regional Development Bureau has sought ways to quantify road image information. If it can be proved that the estimation of road visibility from the same image does not differ among people and if drivers’ assessments of visibility correlate with assessments given by a road image–based visibility scale (VS), then the VS can be used to quantify road visibility. VS is calculated from a weighted intensity of power spectrum developed by Hagiwara and colleagues. Subjective assessments of road images taken by CCTV cameras were surveyed. Assessments of daytime images also had little variation among subjects and correlated closely with the VS; assessments of nighttime images had little variation among subjects but did not correlate closely with the VS. To determine the best way to provide VS information, subjects’ understanding of information given was surveyed in response to three provision media (text, numerical rank, and symbol); the differences were small.
Get full access to this article
View all access options for this article.