基于自适应PCNN的紫外图像融合算法

电晕放电产生的紫外线波长为200~400 nm,为避开太阳光和日光灯的干扰,利用240~280 nm这一日盲波段获取紫外放电图。为能够确定电晕放电的位置,提出了一种适用于可见光图像和紫外图像融合的新算法。以图像空间频率和信息熵作为PCNN对应神经元的链接强度值,并与加权平均法、拉普拉斯金字塔变换方法和小波方法等融合算法相比较。实验结果表明,文中方法优于其它方法,融合效果良好。

UV Image Fusion Based on Adaptive PCNN

The ultraviolet wavelength that corona discharge emits is 200~400 nm.In order to avoid the interference of sunlight and fluorescent lamp,ultraviolet discharge images are acquired by using the sun-blind band from 240 to 280 nm.A visible and UV image fusion algorithm is presented for locating corona discharge.The proposed method uses spatial frequency and information entropy as PCNN linking strength values.Experimental results show that this algorithm offers better fusion results than the weighted average method,the Laplacian-based method and the wavelet-based method.