基于色彩空间和小波分析的多聚焦彩色图像融合

Study on Multi-focus Color Image Fusion Based on Color Space and Wavelet Transform

分别基于RGB、HSV和YIQ三种色彩空间,对多聚焦彩色图像融合进行了大量的研究.在实验中,基于RGB色彩空间时,将图像分解为R分量,G分量,B分量,分别在各个颜色分量上进行小波分解,然后在小波变换域上对所有图像的各个颜色分量的小波分解系数运用加权平均的融合规则,最后做小波反变换得到融合图像.而基于HSV或YIQ色彩空间时,首先要把多聚焦彩色图像转换到HSV或者YIQ色彩空间,然后运用与RGB色彩空间相同的融合方法,最后做小波反变换和色彩空间反变换得到融合图像,实验结果证明,小波分析在多聚焦彩色图像的融合中依然能发挥极好的作用,基于HSV色彩空间的多聚焦彩色图像融合会出现颜色失真,基于RGB和YIQ色彩空间都能取得很好的融合效果,相比之下,基于YIQ色彩空间更能取得较好的融合效果.

In this paper,the multi-focus color image fusion was researched based on RGB,HSV and YIQ spaces.In the experiment.In RGB space,the original images were decomposed into R,G and B color components and each component of the images was decomposed by wavelet.The weighted average fusion rule was applied to wavelet coefficient.The fused image was obtained by the wavelet inverse transform.In HSV or YIQ space,however,the multi-focus color images should be first transformed into HSV or YIQ space,then the same fusion rule with the RGB space was applied.And eventually the fused images were obtained by the wavelet inverse transform and color space inverse transform.The experimental results suggest that the wavelet analysis still play an important role in the multi-focus color image fusion.The multi-focus color image fusion in the HSV color space appears color distortion.Very good fused results were attained in the RGB and YIQ color space.By contrast,the fused results in the YIQ space are better.