摘要
红外和可见光图像的融合既要突出红外图像中重要的亮度特征,又要使融合图像保留清晰的视觉效果。因此,提出了基于Gabor滤波和显著性检测的融合方法。首先,采用显著性检测得到红外和可见光图像的显著层,再使用Frankle-McCann Retinex增强算法对可见光图像进行增强,之后用Gabor滤波器将红外图像和增强后的可见光图像分解为细节层和基础层。然后,采用“最大绝对”的融合策略对显著层与细节层进行融合,最后进行图像重构。实验结果表明,得到的结果与其他八种经典算法比较中表现优异,尤其是AG、EI、IE、SF等指标方面尤为突出。
The fusion of infrared and visible images should highlight the important luminance features in the infrared images,but also make the fused images retain clear visual effects.Therefore,a fusion method based on Gabor filtering and saliency detection is proposed.First,saliency detection is used to obtain the salient layers of the IR and visible images.Then the visible image is enhanced using the Frankle-McCann Retinex enhancement algorithm.After that,the infrared image and the enhanced visible image are decomposed into detail layer and base layer using Gabor filter.Then,a"maximum absolute"fu-sion strategy is used to fuse the significant layer with the detail layer,and finally,the image is reconstructed.The experi-mental results show that the obtained results are superior in comparison with other eight classical algorithms,especially in terms of AG,EI,IE,SF and other indexes.
作者
钟荣军
付芸
ZHONG Rongjun;FU Yun(School of Opto-Electronic Engineering,Changchun University of Science and Technology,Changchun 130022)
出处
《长春理工大学学报(自然科学版)》
2023年第6期26-33,共8页
Journal of Changchun University of Science and Technology(Natural Science Edition)
基金
国家自然科学基金项目(61975021)
吉林省重点科技计划项目(20170204015GX,20180201049YY)。