一种基于HSI和小波变换的可见光和红外图像融合新方法

A new fusion method based on HSI and wavelet transform for visual and infrared images

针对传统的HSI(色度—饱和度—亮度)图像融合方法易出现颜色失真和传统的小波融合方法存在分块模糊的现象,文中提出一种基于HSI和小波变换的可见光和红外图像融合新方法。该方法首先通过HSI变换获得红外和可见光图像的亮度成分,采用基于方差匹配度的融合规则对两者进行小波融合,得到新的亮度分量,并选用红外图像的饱和度分量和可见光图像的色度分量,进行HSI逆变换获得最终融合图像。实验结果表明:该方法在提高图像清晰度、突出图像细节和热目标等方面效果显著。

A new fusion method based on HSI and wavelet transform for infrared and visual images was presented, while the traditional HSI fusion method turned easily out to be color distortion and the traditional wavelet fusion method existed a phenomenon of sub blocks of fuzzy. This method obtained firstly intensity components of infrared and visible images by HSI transform, then used the fusion rule based on the regional variance matching degree to fuse them, so we could get new intensity component. With selecting the saturation component of infrared image and the hue component of visible image, we could obtain the final fused image by HSI inverse transform. The results showed that the proposed method was apparently found to outperform traditional schemes in many cases such as improving the image sharpness, highlighting the image details and hot targets.