摘要
内容复杂且细节丰富的模糊图像,在修复过程中面临难以兼顾噪声去除与细节信息全面保留的问题.为此,提出基于组稀疏残差约束的模糊图像修复方法,在修复图像的同时,最大程度恢复细节特征信息.利用分段线性变换将模糊图像进行灰度变换,利用组稀疏残差约束法去除图像中的噪声,使得去噪后的图像更加接近原始图像.通过构建一个融合注意力机制的增强模型进一步增强细节特征信息,实现模糊图像修复.结果表明:所提方法在PSNR、SSIM和信息熵的平均值上都处于较高水平,说明所提方法可以修复模糊图像样本,修复性能较好.
For fuzzy images with complex content and abundant details,it is difficult to remove noise while fully preserving detailed information during the restoration process.To this end,a fuzzy image inpainting method based on group sparse residual constraint is proposed,which restores detailed feature information to the maximum extent while repairing the image.Using piecewise linear transformation to transform the fuzzy image grayscale,and using group sparse residual constraint method to remove noise from the image,making the denoised image closer to the original image.By constructing an enhanced model that integrates attention mechanisms to further enhance detail feature information,fuzzy image restoration can be achieved.The results show that the proposed method is at a higher level in terms of the average values of PSNR,SSIM,and information entropy,indicating that it can repair fuzzy image samples and has good repairing performance.
作者
陶庆凤
TAO Qingfeng(School of Information Engineering,Minnan Institute of Technology,Shishi 362700,China)
出处
《西安文理学院学报(自然科学版)》
2025年第3期10-15,共6页
Journal of Xi’an University(Natural Science Edition)
基金
福建省中青年教师教育科研项目(科技类)(JAT220433)。
关键词
组稀疏残差约束
模糊图像
细节增强
修复方法
group sparse residual constraint
fuzzy image
detail enhancement
repair method
