面向非均质图像去雾的解耦合三阶段增强网络  

Decoupled triple-stage enhancement network for non-homogeneous image dehazing

作　　者：刘春晓[1] 胡鹏靖 厉世昌 王成骅 凌云[1] Liu Chunxiao;Hu Pengjing;Li Shichang;Wang Chenghua;Ling Yun(School of Computer Science and Technology,Zhejiang Gongshang University,Hangzhou 310018,China)

机构地区：[1]浙江工商大学计算机科学与技术学院,杭州310018

出　　处：《中国图象图形学报》2025年第1期83-94,共12页Journal of Image and Graphics

基　　金：浙江省自然科学基金项目(LY24F020004,LZ23F020004);浙江省重点研发计划项目(2023C01039);浙江工商大学“数字+”学科建设项目(SZJ2022B016);浙江省大学生科技创新活动计划暨新苗人才计划项目(2023R408035,2024R408B080)。

摘　　要：目的在雾霾环境下拍摄的图像通常具有结构对比度较低、细节信息模糊和颜色饱和度失真等特点。虽然目前的去雾算法已经能较好地处理均质雾霾图像,但是对于非均质雾霾图像的去雾能力仍较差。为此,提出了一种面向非均质雾霾图像去雾的解耦合三阶段增强网络。方法通过颜色空间变换将输入图像解耦为亮度、饱和度和色度3个通道之后,该算法首先通过对比度增强模块增强亮度图的对比度,使去雾结果具有更清晰的结构和细节信息;然后,通过饱和度增强模块增强图像的饱和度,使去雾结果具有更鲜艳的颜色;最后,使用颜色矫正增强模块对总体颜色进行微调,使去雾结果更符合人眼视觉感知。特别地,在饱和度增强模块中设计了一个雾霾密度编码矩阵,通过计算亮度图在对比度增强前后的梯度差异,估计出雾霾图像的雾霾密度信息,为饱和度增强模块提供指导,以保证饱和度恢复的准确性。结果在3个数据集上与14种方法进行了对比实验,本文方法在NHD(nonhomogeneous dataset)数据集上得到了最优结果,相比于性能第2的模型,平均峰值信噪比提升了8.5 dB,平均结构相似性提升了0.12;在Real-World数据集中,本文方法的感知雾密度预测值为0.47,雾密度估计值为0.21,均处于前列;在SOTS(synthetic object testing set)数据集中,本文方法的平均峰值信噪比为16.52 dB,平均结构相似性为0.80,在人眼感知效果方面不输于已有方法。结论本文所提方法对于非均质雾霾图像具有优秀的处理能力,可以有效地去除图像的雾霾并还原出雾霾图像的真实细节信息和颜色。Objective The absorption or scattering effect of microscopic particles in the atmosphere,such as aerosols,soot,and haze,will reduce image contrast,blur image details,and cause color distortion.These problems can decrease the accuracy of subsequent advanced computer vision tasks,such as object detection and image segmentation.Therefore,image dehazing has attracted increasing attention,and various image dehazing methods have been proposed.The ultimate goal of image dehazing is to recover a haze-free image from the input hazy image.At present,existing image dehazing algo⁃rithms can be divided into two categories:traditional dehazing algorithms based on image prior and image dehazing algo⁃rithms based on deep learning.The image priori-based dehazing algorithm uses the prior information and empirical rules of the image itself to estimate the transmittance map and atmospheric light value,and it utilizes the atmospheric scattering model to realize the image dehazing process.This approach can improve the contrast of the image to a certain extent but eas⁃ily leads to excessive enhancement or color distortions in the dehazed results.Driven by a large amount of image data,the image dehazing algorithm based on deep learning can flexibly learn the mapping from hazy image to haze-free images by directly constructing an efficient convolutional neural network and obtain dehazed effects with better generalization perfor⁃mance and human visual perception.However,because of domain differences,the image dehazing algorithm trained on the synthesized homogeneous haze dataset usually has difficulty achieving satisfactory results on heterogeneous hazy images in the real world.Method Haze will reduce the contrast of the image and make it look blurry.Thus,we train the network(i.e.,the contrast enhancement module)with the brightness map of the hazy image and the brightness map corresponding to the clear image as the training image pairs,which effectively enhances the contrast of the brightness map and obtains the brightness enhance

关 键 词：深度学习 非均质图像去雾 饱和度增强 对比度增强 三阶段增强 雾霾密度编码矩阵 

分 类 号：T391.41[一般工业技术]