摘要
针对低照度图像增强处理后存在图像细节差、色彩失真和“光晕伪影”等问题,在HSV颜色空间中处理图像,将V通道分量进行小波变换分解,对低频系数采用改进加权引导滤波替代高斯滤波作为环绕函数的多尺度Retinex算法处理,对高频系数采用改进的阈值函数降噪处理,小波逆变换重构V通道分量,对重构的V通道分量进行Gamma校正处理并转换回RGB颜色空间。实验结果表明,该算法增强后图像亮度提升明显,同时抑制了“光晕伪影”,颜色失真较少且细节丰富。
Aiming at the problems of poor image detail,color distortion,and"halo artifacts"after low-illumination image enhancement processing,a low-illumination image enhancement algorithm combining weighted guided filtering and wavelet transform is proposed.The algorithm processes images in HSV color space.This method is used in the HSV color space;the V channel component is decomposed by wavelet transform;the multi-scale Retinex algorithm with the modified guide filtering is used instead of the Gaussian function as the surrounding function for the low frequency coefficient;the improved threshold function noise reduction treatment is used for the high frequency coefficient;the wavelet inverse transformation reconstructs the V channel component;and the reconstructed V channel component is gamma corrected and converted back to the RGB color space.Experimental results show that the image brightness is significantly improved after the algorithm is enhanced,the"halo artifacts"are suppressed,the color distortion is less,and the detail is rich.
作者
王少东
吴春林
刘永城
谢雨晴
王正家
WANG Shaodong;WU Chunlin;LIU Yongcheng;XIE Yuqing;WANG Zhengjia(School of Mechanical Engineering,Hubei Univ.of Tech,Wuhan 430068,China;School of Detroit Greed Tech.Institute,Hubei Univ.of Tech.,Wuhan 430068,China)
出处
《湖北工业大学学报》
2023年第5期14-18,共5页
Journal of Hubei University of Technology
