光子学报, 2019, 48 (1): 0110002, 网络出版: 2019-01-27
基于融合与高斯加权暗通道的单幅图像去雾算法
Single Image Dehazing Algorithm Based on Fusion and Gaussian Weighted Dark Channel
图像融合 高斯权重 图像去雾 暗通道先验 图像复原 Image fusion Gaussian weight Image dehazing Dark channel prior Image restoration
摘要
针对图像去雾算法在景深突变处出现光晕现象和远景区域去雾不足的问题, 提出了一种基于融合与高斯加权暗通道的单幅图像去雾算法.利用图像形态学梯度的特点, 将形态学梯度图像与暗通道图像线性融合获取融合暗通道, 构造自适应的高斯权重参数对融合的暗通道图像逐像素处理获取粗透射率, 在使用L1正则化优化透射率, 通过大气散射模型与修复的大气光值恢复无雾图像.仿真实验表明, 本文算法可以较好地恢复出图像的细节并抑制光晕现象, 与几种典型的图像去雾算法客观对比, 证实了本文算法的可行性.
Abstract
Aiming at the problem that the image dehazing algorithm has halo phenomenon in depth discontinuity and legacy residual fog in the distant area, this paper proposes a single image dehazing algorithm based on fusion and Gaussian weighted dark channel. Firstly, using the characteristics of image morphology gradient, the morphological gradient image and the dark channel image are linearly fused to obtain the fusion dark channel. Secondly, the adaptive Gaussian weight parameter is constructed to pixel-by-pixel process the fused dark channel image to obtain the coarse transmission, and the L1 regularization is used to optimize the transmission. Finally, the haze-free image is restored by the atmospheric scattering model and the restored atmospheric light value. Experimental results show that the proposed algorithm can recover the details of the image and suppress the halo phenomenon. The objective comparison with several typical algorithms proves the feasibility of the proposed algorithm.
张晨, 杨燕. 基于融合与高斯加权暗通道的单幅图像去雾算法[J]. 光子学报, 2019, 48(1): 0110002. ZHANG Chen, YANG Yan. Single Image Dehazing Algorithm Based on Fusion and Gaussian Weighted Dark Channel[J]. ACTA PHOTONICA SINICA, 2019, 48(1): 0110002.