中文摘要：

在雾等恶劣天气备件下，道路交通场景中视线变差，严重影响道路交通安全。因此，如何实现雾天道路交通场景的视觉增强一直是中外相关领域的研究热点。通过对一典型雾天交通场景图像进行分析，采用HSL色彩空间直方图方法、FFT频谱分析方法、小波多尺度子空间能量分布分析等方法，分析并比较了原图、天空区域子图、交通标志子图的特性。提取了反映雾天照明强度和交通标志的有效特征，为天空区域和交通标志的自动分离、雾天交通图像的照明强度的自动估计及雾天交通场景图像的视觉增强和恢复奠定了坚实的基础。

英文摘要：

In bad weather condition, the visibility becomes poor and it may lead traffic accidents. So how to enhance the visibility in foggy weather is one of the hot research topics in correlative scientific area. In this paper, after analyzing a typical foggy traffic image, we compare the character of original image, sky area sub-image, and traffic signs. We extract their features with HSI histogram, FFT spectrum, and wavelet multi-scale methods. These features can be used to separate the sky area from foggy traffic image and enhance the visibility in bad weather.