基于空域分解的雾天图像增强方法  

Fog Image Enhancement Method Based on Spatial Decomposition

作　　者：王鹏[1,2] 齐洁 张鑫 邓守京 孙莉莉 WANG Peng;QI Jie;ZHANG Xin;DENG Shoujing;SUN Lili(Software Technology Institute,Dalian Jiaotong University,Dalian 116028,China;Dalian Portnet Co.,Ltd.,Dalian 116000,China)

机构地区：[1]大连交通大学软件学院,辽宁大连116028 [2]大连口岸物流网有限公司,辽宁大连116000

出　　处：《无线电工程》2021年第11期1261-1269,共9页Radio Engineering

基　　金：人工智能四川省重点实验室开放基金项目(2020RYJ04);辽宁省自然科学基金资助项目(20180551020);辽宁省教育厅资助项目(JDL2019011)。

摘　　要：针对图像去雾问题,提出了一种基于图像的空域分解并与改进的小波阈值去噪算法结合的去雾算法。为了实现图像中噪声和色彩解耦合,将图像采用中值滤波分解为基础层与细节层;进行色彩空间拆分,利用直方图均衡化算法进行基础层的色彩调整,使用改进的小波阈值去噪算法进行细节层噪声抑制;将分层图像进行合成并进行gamma校正,确保去雾后的图像色彩均匀和清晰度提高。验证结果表明,该算法获得的去雾图像细节方面更为丰富,图像清晰度更高。To deal with the problem of image defogging,a defogging algorithm based on image spatial decomposition and combined with improved wavelet threshold denoising algorithm is proposed.Firstly,in order to decouple the noise and color in the image,the image is decomposed into basic layer and detail layer by median filter;then the color space is split,the histogram equalization algorithm is used to adjust the color of the basic layer,and the improved wavelet threshold denoising algorithm is used to suppress the noise of the detail layer;finally,the layered image is synthesized and gamma corrected to ensure the uniformity of image color and the improvement of clarity after defogging.Compared with other algorithms,the experimental results show that the defogged image obtained by this algorithm has richer details and higher image definition.

关 键 词：雾天图像 图像分解 直方图均衡化 小波变换 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]