基于提升小波和直方图均衡化的MSR Retinex多参数融合方法的真彩色图像增强

Color image enhancement based on MSR Retinex parameter fusion of lifting wavelet and histogram equalization

基于原始提升小波和直方图均衡化算法的不足,文章提出了一种基于提升小波的多尺度(MSR)Retinex与直方图均衡化融合的真彩色图像增强算法。该算法将HSV空间下的亮度V分量运用MSR Retinex不同参数所得到的图像提升小波变换,并对低频与高频进行最大值处理,利用提升小波逆变换重构得到新的图像RGB1,并对原始的HSV颜色模型中的亮度分量V进行自适应直方图均衡化处理得到新的图像RGB2,将RGB1与RGB2图像进行融合得到增强后的彩色图像。

As the original lifting wavelet and histogram equalization algorithm are insufficient, a true color image enhancement algorithmbased on the combination of lifting wavelet transform and MSR and histogram equalization fusion is proposed. The algorithm uses MSRRetinex algorithm for luminance component V in HSV space and performs lifting wavelet transform on images obtained by changingdifferent parameters and maximum processing on low frequency and high frequency. Then we reconst the processed low and highfrequency coefficients with lifting wavelet inverse transform to obtain a new image RGB1 and perform adaptive histogram equalization onluminance component V in the original HSV color model to obtain a new image RGB2. Thus, we fuse RGB2 and RGB2 images to obtainan enhanced true color image.