中文摘要：

针对雾天视频图像对比度低、可视性急剧下降的特点,提出一种改进的限制对比度自适应直方图均衡化的视频实时去雾方法。将图像从RGB空间变换到HIS空间,根据模块之间的相似性进行自适应更新,提高图像中心的像素值权重。使用预定义的阈值裁剪直方图,限制噪声的放大和局部对比度的过增强,确保对视频图像处理的快速性。与传统方法进行比较,多组实验结果表明,该方法在信息熵、平均梯度、方差等指标上具有明显的优越性,可以有效去除朦胧,增强雾天图像的层次感;计算复杂度低,可以满足实时性应用的要求。

英文摘要：

For the foggy weather, the video image has the characteristics of low contrast and sharp decline in visual performance, an improved fog video clarity algorithm based on the contrast limited adaptive histogram equalization was proposed for this prob- lem. The input video extracted the RGB components from the image and conversed RGB axes to HIS axes. Then it self-updated adaptively based on the similarity between modules to make the center of the image pixels have a higher weight value. The histo- gram was clipped using a predefined threshold value, which limited the noise amplification and local contrast being over en- hance＆ By these steps, it quickly defogged the fog video images. Comparing with the traditional methods, multiple sets of experimental results show that the presented method has obvious advantages on information entropy, average gradient and variance. It can remove the haze effectively and enhance the layering of fog image. Meanwhile, it has low time complexity, and can meet the requirements of real time applications by performing the real-time dehazing.