基于直觉模糊集的多波段图像融合

摘要

为解决微光探测条件下多波段图像融合效果差、信息丢失的问题,提出了基于直觉模糊集的多波段图像融合方法.首先利用最大化熵准则对源图像分别进行直觉模糊处理,其次对隶属度图像进行去模糊后进行非下采样剪切波变换(NSST),得到各自的高频图像和低频图像,然后对高频图像用绝对值取大融合,对低频图像用清晰度取大融合,最后对高频融合图像和低频融合图像进行NSST 逆变换得到最终融合图像.实验结果表明,本方法可对微光探测条件下多波段图像自适应处理,融合结果目标显著,对比度和清晰度大,主观和客观评价指标均优于其他方法.%To solve the problems of poor fusion effect and loss of information in multi-band images under low-light-level detection conditions, a multi-band image fusion method based on intuitionistic fuzzy set theory is proposed. First, source images are decomposed by intuitionistic fuzzy sets using the maximum entropy criterion. Second, after blurring the membership degree of the image, the image is decomposed by NSST to obtain high- and low-frequency images. Third, the high-frequency images are fused with an absolute value of large, and low-frequency images are fused with a high level of sharpness. Finally, both sets of fusion images are input into the inverse transform of NSST to obtain the final fusion results. The experimental results show that the proposed method can be used to deal adaptively with weak light images. In addition, the results show that the fusion target is remarkable, the contrast and resolution are large, and the subjective and objective evaluation indices are better than those of CNN and DTCWT.