The Determination of the Uniformity in Road Lighting Using Artificial Neural Networks

Abstract

To ensure that drivers can travel safely, it is necessary to provide good visibility conditions of the road lighting. Thanks to good road lighting, accident rates will decrease, pedestrians' safety will be increased and drivers will be able to travel comfortably. Road lighting standards are included in CIE's 115 “Recommendations for the Lighting of Roads for Motor and Pedestrian Traffic”. According to this standard, there are 6 different lighting classes according to the road definition. There are different lighting standards for each class. These are: average luminance (Lave), overall uniformity (U0), longitudinal uniformity (U1), disability glare (TI), lighting of surroundings (SR). Uniformity is a measurement of how equally light is distributed on the road. Overall uniformity ratio is of minimum luminance to mean luminance and longitudinal uniformity is the ratio of minimum luminance to maximum luminance. If the uniformity is good, all objects on the road can be easily seen by drivers. In this study, a new method was used to measure the uniformity of the road. Unlike classical methods, image processing and artificial intelligence techniques are used to calculate luminance and uniformity. The uniformity results of the test roads were examined to meet the standards according to the road class.