基于Retinex的一种图像去雾算法  

Method of Foggy Image Removing Algorithm Based on Retinex Theory

作　　者：林笑君[1] 梁凤梅[1] 

机构地区：[1]太原理工大学信息工程学院,山西太原030024

出　　处：《电视技术》2013年第17期155-158,共4页Video Engineering

基　　金：山西省科技产业化环境建设项目(2010061023)

摘　　要：由于中心环绕Retinex图像增强算法尺度的选择有限,不能在对有雾天气下采集的图像进行有效去雾的同时增强其细节,提出一种基于Retinex算法的曲波变换图像增强算法。首先,根据Retinex算法,先用高斯函数估计出图像的入射分量,再通过Retinex算法将图像的反射分量得出,然后利用曲波变换的多尺度特性将反射分量进行子带分解,对高频子带采用自动变换阈值处理,对低频子带采用线性拉伸,增强其对比度,最后将曲波系数进行合成,得出处理后的图像。实验结果表明,用该方法对雾天图像处理后,图像的纹理细节更清晰,信噪比和信息熵明显提高,视觉效果改善,失真度显著减小。A new curvelet transform image enhancement algorithm based on the Retincx is proposed to improve the center surround Retinex algorithm whose choices is limited about the scale. The proposed algorithm can remove fog effectively while enhance details in the applications of foggy images enhancement. Firstly,the algorithm decomposes the original image into the incidence image and the reflectance image using the Gaussian function. Secondly,the curvelet transform is used to decompose the reflectance into high frequency coefficient and low frequency coefficient. The high frequency coefficient is disposed by auto-commutation threshold. The low frequency coefficient is disposed. At last the curvelet coefficients are reconstructed to obtain the enhancement image. The experimental results show that the proposed method can improve the details, the PSNR, the information entropy and the visual effects. It can decrease the image distortion.

关 键 词：RETINEX算法 曲波变换 USFFT算法 图像去雾 

分 类 号：TN911.73[电子电信—通信与信息系统] TP391.4[电子电信—信息与通信工程]