雾霾天气下基于二次滤波的交通图像去雾算法

雾霾天气下采集到的退化含噪图像模糊不清、对比度较低;而使用传统基于双边滤波的去雾方法得到的图像偏暗,效果有限。针对这些问题,提出了一种新的基于二次滤波的算法,实现雾霾天气下交通图像去雾处理;利用双边滤波对含雾图像的暗通道图像进行第一次滤波,用引导滤波对图像的透射率粗估计进行二次滤波优化。根据降质模型对含雾图像进行复原,进而得到去雾后的图像。实验效果证明,与传统方法相比,得到的去雾图像与真实场景亮度更加相似,色彩饱和度较好,图像质量较高。

The Traffic Image Defogging algorithm based on Twice Filter in Haze Weather

The obtained degraded traffic images with the noise in in haze weather have low contrast. The result of the traditional bilateral filtering algorithm is too dark to be suitable for detecting. To deal with these problems, the traffic image defogging algorithm based on Twice Filter in Haze Weather was proposed. The first processing is bilateral filtering the dark-channel image of the degraded traffic image. And the Guide filter are used to optimize the rough estimation of image transmittance. Image restoration is realized by the degradation model recovery and defogging image is obtained. Experimental results prove that the obtained defogging image in the proposed algorithm is more similar to the real scene brightness. The color saturation is better and image quality is relatively high.