多聚焦图像焦深扩展分解融合算法仿真  

Simulation of Multi Focus Image Focal Depth Expansion Decomposition Fusion Algorithm

作　　者：刘丽丽[1] 赵玉 LIU Li-li;ZHAO Yu(Computer Department,Changzhi University,Changzhi Shanxi 046011,China;School of Civil Engineering Shijiazhuang Tiedao University,Shijiazhuang Hebei 050043,China)

机构地区：[1]长治学院计算机系,山西长治046011 [2]石家庄铁道大学土木工程学院,河北石家庄050043

出　　处：《计算机仿真》2023年第4期195-198,218,共5页Computer Simulation

摘　　要：多聚焦图像在融合过程中,若源图像分辨率过低会直接影响图像的融合效果。为提升图像融合效果,提出基于焦深扩展的多聚焦图像融合算法。采用焦深扩展方法对光学系统实施优化,通过优化后的光学系统生成聚焦图像,并对获取的聚焦图像中的退化图像实施修复处理,提升源聚焦图像质量;使用双三次插值的单帧超分辨率法对图像实施分解处理,获取源聚焦图像的高分辨率图像,结合滤波处理算法,获取聚焦图像的细节系数;利用逆小波变换方法对图像细节系数,依据计算结果实现多聚焦图像融合。实验结果表明,使用上述方法开展图像融合时,融合效果较好,图像融合质量较高。In order to improve the fusion effect of images,a multi-focus image fusion algorithm based on focal depth extension was proposed.At first,the focal depth expansion method was adopted to optimize the optical system,and then the focused image was generated by the system.Meanwhile,the degraded image in the focused image was repaired to improve the quality of the source image.Moreover,the single-frame super-resolution method based on bicubic interpolation was adopted to decompose the image and thus obtain the high-resolution image of the source image.Combined with the filtering algorithm,the detail coefficient of the focused image could be obtained.Finally,the inverse wavelet transform method was used to calculate the detail coefficients of the image.Thus,multi-focus image fusion was achieved.Experimental results show that the proposed method has better fusion effect and fusion quality.

关 键 词：焦深扩展 多聚焦图像 光学系统优化 图像分解 图像融合算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]