摘要
现有基于小波变换的增强方法对雾天图像增强效果一般,对此提出一种结合小波变换和改进的Retniex的雾天图像增强方法。使用小波变换将图像分解为高频部分和低频部分;利用双边滤波良好的边缘保持性,对传统的Retinex算法进行改进。用改进后的Retinex算法对低频子带图像进行去噪处理。为了防止放大噪声、过度增强,用模糊规则增强算法对高频子带图像进行增强,进行小波逆变换,得到增强后的雾天图像。实验结果表明,基于小波变换和改进的Retniex方法对雾天图像的增强有显著效果。
Since the enhancement method based on wavelet transform has a general effect on foggy images,a Retniex image enhancement method based on wavelet transform and improved Retniex is proposed.The wavelet transform was used to decompose the image into high frequency part and low frequency part;the traditional Retinex algorithm was improved by using the good edge retention of bilateral filtering.The improved Retinex algorithm was used to denoise the low-frequency sub-band images.In order to prevent noise amplification and excessive enhancement,fuzzy rule enhancement algorithm was applied to enhance the low-frequency sub-band image,and then did inverse wavelet transform to get the enhanced foggy images.The experimental results show that the Retniex method based on wavelet transform and improved Retniex is effective in enhancing foggy images.
作者
张振华
陆金桂
Zhang Zhenhua;Lu Jingui(School of Mechanical and Dynamic Engineering,Nanjing Tech University,Nanjing 211816,Jiangsu,China)
出处
《计算机应用与软件》
北大核心
2021年第1期227-231,共5页
Computer Applications and Software
