基于改进拉普拉斯金字塔的红外图像增强算法
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

2022年度黑龙江省省属高等学校基本科研业务费科研项目(No.2022-KYYWF-0527)资助。


Infrared image enhancement algorithm based on improved Laplacian pyramid
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:

    针对在进行红外图像增强时容易出现细节和边缘纹理丢失的问题,提出了基于改进拉普拉斯金字塔的红外图像增强算法。首先,构建拉普拉斯金字塔时,在原有的差分运算中加入Canny边缘检测,提取图像的基础层和细节层;其次,在基础层使用γ CLAHE算法改善对比度和亮度;对细节层通过拉普拉斯算子进一步增强细节层中的边缘纹理;最后将细节层与基础层重建得到增强后的红外图像。经实验验证,本算法与传统Clahe算法、Gamma校正及其他算法相比,其中,PSNR最大提高了534,SSIM值最大提高了06,熵值最大提高了207,验证了本算法能够在红外图像增强时提高对比度,突出边缘信息,保持结构特性完整,在红外图像增强处理中是有效的。

    Abstract:

    In this paper,an infrared image enhancement algorithm based on an improved Laplacian pyramid is proposed to address the common issue of detail and edge texture loss during infrared image enhancement.Firstly,in constructing the Laplacian pyramid,Canny edge detection is incorporated into the existing difference operation to extract the base and detail layers of the image.Secondly,the γ CLAHE algorithm is applied to improve contrast and brightness in the base layer,and the Laplacian operator is then used to further enhance edge textures in the detail layer.Lastly,the enhanced infrared image is reconstructed by combining the detail layer with the base layer.Experimental results demonstrate that compared to traditional methods such as Clahe algorithm,Gamma correction,and others,the proposed algorithm achieves a maximum increase of 5.34 in PSNR,0.6 in SSIM,and 2.07 in entropy,which validates the algorithm′s effectiveness in enhancing contrast,highlighting edge information,and preserving structural characteristics during infrared image enhancement.

    参考文献
    相似文献
    引证文献
引用本文

韩龙,赵雅婷,左超,何辉煌.基于改进拉普拉斯金字塔的红外图像增强算法[J].激光与红外,2024,54(10):1626~1632
HAN Long, ZHAO Ya-ting, ZUO Chao, HE Hui-huang. Infrared image enhancement algorithm based on improved Laplacian pyramid[J]. LASER & INFRARED,2024,54(10):1626~1632

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: 2024-10-16
  • 出版日期: