计算机科学与探索 ›› 2022, Vol. 16 ›› Issue (2): 448-457.DOI: 10.3778/j.issn.1673-9418.2009009

• 图形图像 • 上一篇    下一篇

高光谱协同稀疏与非局部低秩张量变化检测

詹天明1,2, 宋博1,+(), 孙乐3, 万鸣华1, 杨国为1   

  1. 1.南京审计大学 信息工程学院,南京 211815
    2.南京审计大学 审计信息工程与技术协同创新中心,南京 211815
    3.南京信息工程大学 计算机学院,南京 210094
  • 收稿日期:2020-09-04 修回日期:2020-11-20 出版日期:2022-02-01 发布日期:2020-12-09
  • 通讯作者: + E-mail: mg1909003@stu.nau.edu.cn
  • 作者简介:詹天明(1984—),男,江苏高邮人,博士,副教授,硕士生导师,主要研究方向为高光谱图像处理、深度学习。
    宋博(1997—),男,江苏南京人,硕士研究生,主要研究方向为高光谱图像处理。
    孙乐(1987—),男,江苏洋河人,博士,副教授,硕士生导师,主要研究方向为高光谱图像 处理。
    万鸣华(1978—),男,江西南昌人,博士,副教授,硕士生导师,主要研究方向为模式识别、数据处理。
    杨国为(1964—),男,江西樟树人,博士,教授,硕士生导师,主要研究方向为人工智能、大 数据。
  • 基金资助:
    国家自然科学基金面上项目(61976117);国家自然科学基金面上项目(61876213);江苏省自然科学基金面上项目(BK20191409);江苏省高校自然科学研究重大项目(19KJA360001);江苏省高校自然科学研究重大项目(18KJA520005);审计信息工程与技术协同创新中心2018年度研究课题(18CICA09);南京审计大学青年教师科研培育项目(18QNPY015);江苏省研究生科研与实践创新计划项目(KYCX20_1680)

Hyperspectral Change Detection Using Collaborative Sparsity and Nonlocal Low-Rank Tensor

ZHAN Tianming1,2, SONG Bo1,+(), SUN Le3, WAN Minghua1, YANG Guowei1   

  1. 1. School of Information Engineering, Nanjing Audit University, Nanjing 211815, China
    2. Collaborative Innovation Center of Audit Information Engineering and Technology, Nanjing Audit University, Nanjing 211815, China
    3. School of Computer Science, Nanjing University of Information Science and Technology, Nanjing 210094, China
  • Received:2020-09-04 Revised:2020-11-20 Online:2022-02-01 Published:2020-12-09
  • About author:ZHAN Tianming, born in 1984, Ph.D., associate professor, M.S. supervisor. His research interests include hyperspectral image processing and deep learning.
    SONG Bo, born in 1997, M.S. candidate. His research interest is hyperspectral image processing.
    SUN Le, born in 1987, Ph.D., associate professor, M.S. supervisor. His research interest is hyper-spectral image processing.
    WAN Minghua, born in 1978, Ph.D., associate professor, M.S. supervisor. His research interests include pattern recognition and data processing.
    YANG Guowei, born in 1964, Ph.D., professor, M.S. supervisor. His research interests include artificial intelligence and big data.
  • Supported by:
    National Natural Science Foundation of China(61976117);National Natural Science Foundation of China(61876213);Natural Science Foundation of Jiangsu Province(BK20191409);Key Projects of University Natural Science Fund of Jiangsu Province(19KJA360001);Key Projects of University Natural Science Fund of Jiangsu Province(18KJA520005);Project of Collaborative Innovation Center of Audit Information Engineering and Technology(18CICA09);Cultivation Project of Nanjing Audit University(18QNPY015);Postgraduate Research and Practice Innovation Program of Jiangsu Province(KYCX20_1680)

摘要:

高光谱图像变化检测可提供地球表面的时间维变化信息,对城乡规划和管理至关重要。因具有较高的光谱分辨率,高光谱图像常被用于检测更精细的变化。针对高光谱变化检测的问题,提出一种基于协同稀疏与非局部低秩张量的高光谱图像变化检测方法。该方法首先求得前后时间点的高光谱差分图像,再根据差分图像中图像块的非局部分布特点,提取不同的非局部张量簇。然后基于协同稀疏正则化和低秩正则化建立协同稀疏与非局部低秩张量变化检测模型,并采用交替方向乘子法对模型求解得到表示系数。最后根据表示系数求得张量在不同类别中的投影残差,进而根据投影残差最小准则判断该张量块是否发生变化。在Farm-land数据集和Urban area in San Francisco City数据集上进行实验,实验结果表明该方法取得较好的高光谱变化检测精度。

关键词: 高光谱, 变化检测, 协同稀疏, 非局部低秩, 张量分解

Abstract:

Hyperspectral image change detection can provide timely change information on the surface of the earth, which is essential for urban and rural planning and management. Due to the higher spectral resolution, hyperspectral images are often used to detect finer changes. Aiming at the problem of change detection by using hyperspectral image, a hyperspectral change detection method based on collaborative sparsity and nonlocal low-rank tensor is proposed. This method first obtains hyperspectral differential image at different time points, and then extracts different nonlocal similar block tensor clusters according to the nonlocal distribution characteristics of the image blocks in the differential image. Then, based on collaborative sparse regularization and low-rank regularization, a change detection model using collaborative sparsity and non-local low-rank tensor is established, and the representa-tion coefficient is obtained by solving the model using the alternating direction method of multipliers. Finally, the projection residuals of the tensor in different categories are obtained according to the representation coefficients, and then the projection residual minimization criterion is judged whether the tensor has changed. Experiments on Farm-land and Urban area in San Francisco City datasets demonstrate that the proposed method can achieve much better changes detection accuracy.

Key words: hyperspectral, change detection, collaborative sparsity, non-local low-rank, tensor decomposition

中图分类号: