基于容差机制与高斯滤波的图像去雾算法  

Research on Image Defogging Algorithm Based on Tolerance Mechanism and Gaussian Filtering

作　　者：石冬阳 张俊林 刘天光 武正萍 SHI Dongyang;ZHANG Junlin;LIU Tianguang;WU Zhengping(College of Electrical Engineering,Chongqing University of Science and Technology,Chongqing,401331,China)

机构地区：[1]重庆科技学院电气工程学院,重庆401331

出　　处：《重庆科技学院学报（自然科学版）》2023年第5期56-62,共7页Journal of Chongqing University of Science and Technology：Natural Sciences Edition

基　　金：国家级大学生科技创新训练计划项目“基于图像处理的输电线路绝缘子自爆故障诊断方法研究”(202111551009);重庆市自然科学基金面上项目“基于操作行为的驾驶人疲劳特征学习方法研究”(CSTC2020JCYJ-MSXMX0927);重庆科技学院硕士研究生创新计划项目“基于Mask RCNN的自动驾驶系统目标检测方法研究”(YKJCX2220415)。

摘　　要：为了使暗通道先验去雾算法能够有效处理雾霾图像中天空等明亮区域,提升去雾后图像的抗噪性能,提出了一种基于容差机制与高斯滤波的图像去雾算法。首先,引入容差机制来修正明亮区域透射率;然后,融入高斯滤波算法来提升去雾后图像的抗噪性能;最后,适当调整图像亮度,以提升去雾后图像的可视化效果。仿真实验结果表明,改进后算法不仅能够有效处理雾霾图像中天空等明亮区域,而且能够提升去雾后图像的抗噪性能;相较于主流的同态滤波算法,改进后算法的去雾性能更优。In order to make the dark channel prior defogging algorithm effectively deal with bright areas such as the sky and improve the anti-noise performance of the defogged image,an image defogging algorithm based on tolerance mechanism and Gaussian filtering is proposed.Firstly,a tolerance mechanism is introduced to correct the transmittance of bright region.Secondly,gaussian filtering algorithm is integrated into the above defogging algorithms to improve the anti-noise performance of the defogged image.Finally,the brightness of the image is adjusted appropriately to improve the visualization effect of the image after removing the fog.Simulation results show that the improved defogging algorithm can not only effectively deal with bright areas such as the sky in the haze image,but also improve the anti-noise performance of the defogged image.At the same time,compared with mainstream algorithms such as homomorphic filtering,the improved defogging algorithm can obtain better defogging performance,which verifies the effectiveness and superiority of the proposed algorithm.

关 键 词：暗通道先验 抗噪性能 容差机制 高斯滤波 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]