Automated Mobile Highway Sign Visibility Measurement System

This paper describes a prototype of a digital video image analysis system that measures the relative luminance of highway signs using nighttime video images taken by a camera mounted in a moving vehicle. The concept is that by using an inexpensive video camera and laptop computer mounted in a vehicle a measurement of the brightness and visibility of the sign can be made. The system uses real time image analysis of night images of signs acquired on a mobile platform moving at highway speeds, illuminated by normal vehicle head lamps. Using video images at night results in measurements that as closely as possible reflect what the human eye sees. This is an improvement over current practice for ensuring that the signs are performing at night as required. Current methods consist of either visual inspection, measurement of retroreflectivity with a hand-held retroreflectometer, or systematic periodic replacement of signs irregardless of whether they are deficient or not. None of these approaches are ideal. Visual inspection is too subjective, hand-held retroreflectometer measurements are tedious, time-consuming, and sometimes dangerous, and systematic replacement is not cost effective. Results of this research have not only demonstrated that this is a viable way to evaluate the effectiveness of the signs at night, but research has also shown that the retroreflectivity of signs as measured by portable retroreflectometers is not well correlated with the visibility of the signs.