Calibration and use of camera-based systems for road lighting assessment

Standard methods of road lighting quality evaluation are contrasted with the availability of automated measurement systems. A dynamic system based on a CCD camera on a moving vehicle is then described. Determination of the camera position and orientation is required in order to relate the pixel information to the road luminance standards. The image-processing techniques of image flow and feature-based methods are used to identify position, along with road edge following. Methods are derived that permit curved roads to be followed and luminance values to be estimated. A comparison between a luminance map obtained by manual measurement and automated measurement of a physical lighting installation is presented.