Adverse weather such as fog will reduce drivers' visual recognition ability, and important visual information in the driving will be weakened and even lost, which brings risks to driving safety. This paper sets out to explore causes for the decline in driving visual ability in foggy sections at night based on drivers' visual recognition demand mechanism. The purpose is to study the influence of light source characteristics on visual distance under open traffic conditions. Through an outdoor visual recognition test, data were collected on the visual recognition distances of 12 car drivers to standard gray small targets, under different light source characteristics and meteorological visibility conditions. Results revealed that the latter conditions, along with illuminance and correlated color temperature affected drivers' visual recognition ability in low meteorological visibility sections at night. At the same illuminance and meteorological visibility, a light source with high correlated color temperature could improve visual distance. The multivariate power function model of meteorological visibility, illuminance at the target and driver's visual distance in foggy areas were obtained by fitting. The research methods and conclusions could provide technical reference for improving traffic visual distance at night, and thus help lay a foundation for road construction and traffic control technology in foggy areas.