Optimization of tunnel entrance lighting under dynamic visual conditions based on perceived contrast threshold and recognition time analysis

Abstract

Accurate determination of perceived contrast threshold $C_{min}$ is essential for optimizing the design and evaluation of tunnel entrance lighting. However, conventional methods often fail to adequately capture the $C_{min}$ requirements under dynamic driving conditions. This study analyses the interactive effects of tunnel external luminance ( $L_{\mathrm{e}}$ ) and tunnel entrance road surface luminance (L_b) on drivers’ obstacle recognition ability. Under both positive and negative luminance contrast conditions, the $C_{min}$ required to achieve reliable recognition (80% perception probability) and the associated recognition time ( $T_{\mathrm{r}}$ ) were determined across various luminance combinations. The results indicate that compared to the result from the fixed value stimulation involving static visual targets, the $C_{min}$ required for drivers to reliably recognize a fixed obstacle under simulated dynamic driving conditions increased by up to 35% ( $L_{\mathrm{e}}$  = 4000 cd m⁻², $L_{\mathrm{b}}$  = 75.8 cd m⁻²). The required $L_{\mathrm{b}}$ rises in a nonlinear and accelerated manner as $L_{\mathrm{e}}$ increases, to maintain the same visual recognition efficacy. Visual recognition performance under negative luminance contrast conditions is significantly better than under positive contrast, which is characterized by smaller $C_{min}$ and shorter $T_{\mathrm{r}}$ . Luminance combination models for general safety and high safety recognition time are established. A comparison with existing standards revealed the limitations of current specifications for tunnel entrance luminance design.