In order to implement the recommended Commission Internationale de l'Eclairage (CIE) system for mesopic photometry to roads, it is necessary to define the relevant visual field and adaptation luminance in night-time driving conditions. We measured three drivers' eye tracking on a rural road at night and in daytime, and the simultaneous luminance for the corresponding parts of the scene on lit and unlit sections of the road at night. Fields of view with circular sizes of 18, 58, 108, 158 and 208, with the centre point at the mode of the gaze distributions of the drivers, were used as initial estimates of the visual adaptation field. In both the lit and unlit sections, the variation within subject and between subjects in the mean luminance decreased as the size of the circular field increased. However, the mean luminances of all of the circular fields in the unlit section were higher than in the lit section due to the use of high-beam headlights in the unlit section.

The definition of the visual adaptation field in night-time driving conditions is important to implement the CIE system for mesopic photometry. For this purpose, reaction times to an achromatic circular stimulus appearing at different horizontal and vertical eccentricities were measured on uniform and non-uniform backgrounds under mesopic light levels. A large screen illuminated by three projectors providing a visual field subtending 1808 Â 448 was used to show the background and stimuli. The results concerning both the uniform and non-uniform fields indicate that reaction times under mesopic light levels depend on the local luminance of the visual task. The location and the luminance of the stimulus and the corresponding background luminance at that point strongly affect the reaction times in mesopic conditions. Luminance distribution of the non-uniform backgrounds also affects visual performance.

Properly defining the visual adaptation field in night-time driving conditions is important in order to implement the CIE 191 system of mesopic photometry. In night-time driving, visual scenes include several objects with different luminances. The luminance distribution affects visual performance. In this study, contrast threshold measurements were conducted using simulated night-time driving images taken from a street illuminated by LED light sources. A large screen illuminated by three projectors, providing a visual field subtending 1808 Â448, was used to provide the background and target. Ten subjects (mean age 30 years) participated in the experiment. The subjects were asked to detect the targets appearing at different eccentricities (from À608 to 608). The results indicate that the target luminance and luminance distribution of the surrounding area affects peripheral target detection. Under similar luminance distributions, the contrast threshold values did not differ between circular 108 and 208 fields of view. The results support the importance of the area surrounding the road for the visual adaptation field.

No abstract