基于点扩散函数的机载光电吊舱图像增强融合

Image enhanced fusion of photoelectric pod based on point spread function

为更好地融合机载光电吊舱中多源图像的特征信息,提出一种利用点扩散函数(PSF)对多源图像进行增强融合的方法。采用刃边法和小波变换法分别对红外和可见光图像进行点扩散函数求取,利用维纳滤波恢复得到增强图像,然后采用高斯和双边联合滤波对图像进行多尺度混合分解,针对不同尺度下的分解图像,在确定融合权重时引入红外和可见光图像点扩散函数的半幅全宽(FWHM)比值作为分辨率补偿因子,以获得更高质量的融合图像。实验结果表明通过引入点扩散函数信息,光电吊舱系统的获得的融合图像能够具有更好的对比度和分辨率信息,主观和客观评价结果都有所改善。

In order to better integrate the feature information of multi-source images in airborne photoelectric pods,a method of enhancing fusion of multi-source images using point spread function(PSF)is proposed.For the infrared and visible images,knife-edge method and wavelet transform method are used to obtain the point spread function.Wiener filter is used to recover the enhanced image.Gaussian and bilateral joint filtering are used to realize multi-scale hybrid decomposition of the image.For the decomposition images at different scales,the full width at half maxima(FWHM)ratio of the infrared and visible image PSF is introduced as a resolution compensation factor when determining the fusion weights to obtain a higher quality fused image.The experimental results show that by introducing the point spread function information,the fused image can have better contrast and resolution information,and the subjective and objective evaluation results are both improved.