Visibility Estimation Using Roadside Cameras: A Calibration Based Method

Poor visibility in bad weather has significant impacts on traffic safety. An ability to accurately monitor visibility can help reduce the driving risk under this condition. However, visibility measuring based on sensors are precise but costly, and measuring based on vision are lower cost but more complicated in training and deployment. This paper proposes a simple yet effective method to estimate visibility using roadside cameras. The method is established based on the definition of visibility in meteorology, taking a great distance at which an object can be clearly discerned from a bright background as the visibility. The method takes the sky region as the reference background image and using relationship mapping for the distance measurements. This proposed method simplifies the calibrating and measuring process, visibility estimation can be obtained without training or configuration, making it easier to use in practice.