基于多尺度引导滤波的多光谱图像去模糊方法研究
DOI:
CSTR:
作者:
作者单位:

中国电子科技集团公司第41研究所

作者简介:

通讯作者:

中图分类号:

TP391

基金项目:

安徽省自然科学(2308085QF215);国家自然科学(62127804);


Research on Multi-Scale Guided Filtering Algorithm for Multispectral Image Deblurring
Author:
Affiliation:

1.Science and Technology on Electronic Test Measurement Laboratory,The 41th Research Institute of China Electronics Technology Group Corporation;2.The 41th Research Institute of China Electronics Technology Group Corporation

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:

    多光谱图像的在计算机视觉任务中发挥着重要的作用。针对可见光-近红外波段多光谱成像系统的跨通道图像模糊的问题,提出了一种基于多尺度引导滤波的多光谱图像去模糊方法。该方法建立了一个多尺度引导滤波的图像去模糊框架,将清晰图像作为先验图像,利用金字塔形式逐渐恢复点扩散函数的最优解,并将梯度先验约束条件带入到最大后验概率框架实现模糊图像复原。实验表明,本文提出的算法解决了直接在高分辨率图像估计模糊核偏差较大的问题,能够对多光谱图像集中的模糊图像进行快速复原,相比其他对比方法重建得到的图像具有出色的纹理细节与边缘保持效果,在无参考感知模糊指标上具有显著优势。该方法对于多光谱图像跨通道信息融合与计算光学成像系统性能跃升具有重要的实用价值。

    Abstract:

    Multispectral images play a vital role in computer vision tasks. To address the cross-channel image blur problem in visible-near-infrared band multispectral imaging systems, this paper proposes a multispectral image deblurring method based on multi-scale guided filtering. A multi-scale guided filtering framework for image deblurring is established, utilizing clear images as prior references. Through a pyramid structure, it progressively restores the optimal solution of the point spread function and blurred image restoration with incorporates gradient prior constraints into a maximum a posteriori probability framework. Experimental results demonstrate that the proposed algorithm effectively resolves the issue of significant deviations in direct high-resolution image blur kernel estimation. It enables rapid restoration of blurred images in multispectral image sets, outperforming other comparative methods in terms of reconstructed image quality with superior texture details and edge preservation. The method exhibits a significant advantage in the evaluator metric. This approach provides valuable insights for cross-modal information fusion in multispectral imaging and the performance enhancement of computational optical imaging.

    参考文献
    相似文献
    引证文献
引用本文
分享
相关视频

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:2025-07-05
  • 最后修改日期:2025-10-27
  • 录用日期:2025-11-07
  • 在线发布日期:
  • 出版日期:
文章二维码