基于分数混合阶的交叠组合稀疏全变分 HFOGSTV 去噪模型
DOI:
CSTR:
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TN911;TP751.1

基金项目:

国家级大学生创新创业训练计划支持项目(202310300036Z)、南京信息工程大学大学生创新创业训练计划项目 (XJDC202310300326)资助


Overlapping combined sparse total variation HFOGSTV denoising model based on fractional mixing order
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对全变分(TV) 模型引起的阶梯效应,造成图像出现块状伪影的问题,提出了基于分数混合阶的交叠组合稀疏正则 项全变分(HFOGSTV) 去噪模型。考虑到交叠组合稀疏正则项与分数阶微分能够抑制全局以及局部的块状伪影的特点,将二 者引入 TV 模型,可以较好地抑制阶梯效应。并引入高阶微分,提高模型的去噪效率,构建了基于 HFOGSTV 去噪模型。新 模型采用交替方向乘子法(ADMM)分解为各个子问题进行依次迭代求解,并选择合适的参数。实验结果表明,HFOGSTV 模 型与非局部均值去噪(NLM)、TV 和混合非凸高所交叠组合稀疏全变分(HNHOTV-OGS) 模型相比,峰值信噪比(PSNR) 分 别提升7.2%、5.3%、1.9%,结构相似性(SSIM)分别提升6.6%、6.1%、3.4%,耗时分别减少89%、51%、45%,不仅有效抑制 了阶梯效应,而且去噪效果更佳,运行时长大大降低。

    Abstract:

    Aiming at the problem of block-like artifacts in images caused by the staircase effect caused by total variation (TV)model,this paper proposes an overlapping combined sparse regular term denoising HFOGSTV model based on fractional mixing order.Considering that the overlapping combination of sparse regularization term and fractional differentiation can suppress global and local block artifacts,this paper introduces them into the TV model to better suppress the ladder effect.Furthermore,high-order differentiation is introduced to improve the denoising efficiency of the model,and an overlapping combined sparse regular term denoising HFOGSTV model based on fractional mixing order is constructed.The new model is decomposed into each subproblem by alternating direction multiplier ADMM method,and the appropriate parameters are selected.The experimental results show that compared with the NLM, TV,and HNHOTV-OGS models,the PSNR of the HFOGSTV model increases by 7.2%,5.3%,and 1.9%, respectively,and the SSIM increases by 6.6%,6.1%,and 3.4%,respectively,while the running time decreases by 89%,51%,and 45%,respectively.It not only effectively suppresses the staircase effect,but also has better denoising effect and greatly reduces the running time.

    参考文献
    相似文献
    引证文献
引用本文

杨传兵,周先春,陈 楷,张 洁.基于分数混合阶的交叠组合稀疏全变分 HFOGSTV 去噪模型[J].国外电子测量技术,2024,43(2):74-83

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2024-05-29
  • 出版日期:
文章二维码