中文摘要：

针对因大气颗粒吸收及散射因素带来视频图像质量下降的问题，分析并结合传统的基于暗通道先验的规律图像去雾算法的基础上，提出一种基于通道暗元素信息的快速单幅图像去雾方法。该方法通过计算出单像素最小分量矩阵，然后复用该单像素最小分量矩阵计算出其余滑窗模板下图像对应的最小分量矩阵，最后结合图像的边缘信息选择不同滑窗模板对应的场景透射系数完成对图像的去雾。实验与传统图像去雾算法作比较表明，该方法恢复出的图像质量较高，解决了图像去雾处理时物体边缘附近易产生光晕的问题；避免了原始算法的软抠图处理，计算量小，能够达到实时处理的目的，便于实际应用。

英文摘要：

To effectively solve the problem of video image quality degradation caused by the absorption and scattering of atmospheric particles, through analyzing and combining with traditional image dehazing method based on the prior rule of dark channel, this paper proposed a fast signal image dehazing method based on channel dark element information.The basic idea of this method was calculating a single pixel minimum component matrix.Then it multiplexed the signal pixel minimum component matrix to parallel calculate other sliding window template under image corresponding to the minimum weight matrix.Finally, according to the edge information of the image, selected the transmission coefficients of the different sliding window templates which to complete the image dehazing.The experimental of comparing this method with the traditional image dehazing method, which verifies this method to restore the image of high quality and solving the problem of easy to produce a halo near the edges of object, moreover avoiding the original method of soft matting, small amount of calculation achieving the purpose of real-time processing and convenient in practical application.