|
Abstract
|
Objective: Specifications for lighting systems used in the driving environment are based on limited empirical experimental evidence as well as the consensus of practitioners and experts involved in standards-making bodies. The extent to which factors such as driver age or visual function are considered in final versions of such specifications might often only be understood qualitatively. Several models for characterizing visual performance for specific visual tasks and by different populations exist, but little is known about if and how well such models can be applied to conditions experienced while driving.
Methods: Two models for characterizing visual performance, one for on-axis visibility (Rea and Ouellette, 1991) and an- other for peripheral visual detection (Bullough and Van Derlofske, 2004), were used to generate predictions of the speed and accuracy of visual processing under conditions utilized in several laboratory and field studies pertaining to lighting and visual information such as signage in transportation applications.
Results: When applied appropriately (e.g., using on-axis models for on-axis visual tasks, and off-axis models for off-axis tasks), published computational models such as those of on- and off-axis visual performance can work well for predicting visual performance under a wide range of viewing conditions and for different populations (such as different age groups). Strong correlations between modeled and observed visual performance were found in a number of studies with diverse experimental methods.
Conclusions: The results bode well for the systematic use of models such as those investigated in the present paper for evaluating the impact of different lighting alternatives on visibility for a wide array of visual tasks. Specifications for lighting systems used in the driving environment could perhaps be made in terms of visual performance rather than in terms of light level as is presently the case. When empirical experimental evidence is still required to justify a particular lighting approach, the modeling techniques described here will help to formulate efficient testing of hypotheses. |