Images of roads captured by visual surveillance system are usually degraded by scattering due to atmospheric particles such as haze, fog and mist, which could frequently bring great difficulties to the transportation. In this paper, a novel method was proposed based on weighted least squares(WLS) to remove fog from a single input image. The proposed algorithm begins with estimation of atmospheric light and white balance. Then, through the constraint from the road environment, the weighted least squares(WLS) framework was constructed to estimate atmospheric veil, and restore the fog image by inverting the atmospheric scattering model. The experimental results demonstrates that, compared with the existing algorithm, the proposed method can remove the fog effectively, eliminate the Halo effects, obtain good restoration of distant scene details, and thus realize the improvement of traffic image visibility.