电子学报 ›› 2020, Vol. 48 ›› Issue (12): 2453-2461.DOI: 10.3969/j.issn.0372-2112.2020.12.022

• 学术论文 • 上一篇    下一篇

光谱加权协同稀疏和全变差正则化高光谱图像解混

张绍泉, 黄志浩, 邓承志, 李璠, 徐晨光, 吴朝明, 汪胜前   

  1. 南昌工程学院江西省水信息协同感知与智能处理重点实验室, 江西南昌 330099
  • 收稿日期:2019-10-21 修回日期:2020-07-09 出版日期:2020-12-25
    • 通讯作者:
    • 邓承志
    • 作者简介:
    • 张绍泉 男,1990年生于江西抚州,博士,讲师,主要研究方向为高光谱遥感图像处理,机器学习等研究.E-mail:zhangshaoquan1@163.com;黄志浩 女,1995年生于云南曲靖,硕士研究生,主要研究方向为高光谱图像解混.E-mail:huangzhihao1119@163.com;李璠 女,1983年生于湖北襄阳,硕士,讲师,主要研究方向为高光谱图像解混、机器学习等.E-mail:lifan@nit.edu.cn;徐晨光 男,1985年生于江西抚州,硕士,讲师,主要研究方向为高光谱图像解混、机器学习等.E-mail:xcg@nit.edu.cn;吴朝明 男,1979年生于江西抚州,博士,讲师,主要研究方向为图像处理、视觉伺服,高光谱遥感图像处理.E-mail:zmwunit@foxmail.com;汪胜前 男,1965年生于江西浮梁,博士,教授,主要研究方向为图像处理,信息安全等.E-mail:Sqwang113@263.com
    • 基金资助:
    • 江西省教育厅科技项目 (No.GJJ180962,No.GJJ190956); 国家自然科学基金 (No.61901208,No.61865012); 江西省自然科学基金 (No.20192BAB217003,No.20181ACG70022); 南昌工程学院2019年度研究生创新计划项目 (No.YJSCX20190014)

Spectral Reweighted Collaborative Sparsity and Total Variation Based Hyperspectral Unmixing Method

ZHANG Shao-quan, HUANG Zhi-hao, DENG Cheng-zhi, LI Fan, XU Chen-guang, WU Zhao-ming, WANG Sheng-qian   

  1. Jiangxi Province Key Laboratory of Water Information Cooperative Sensing and Intelligent Processing, Nanchang Institute of Technology, Nanchang, Jiangxi 330099, China
  • Received:2019-10-21 Revised:2020-07-09 Online:2020-12-25 Published:2020-12-25
    • Corresponding author:
    • DENG Cheng-zhi
    • Supported by:
    • Science and Technology Project of Education Department of Jiangxi Province (No.GJJ180962, No.GJJ190956); National Natural Science Foundation of China (No.61901208, No.61865012); Natural Science Foundation of Jiangxi Province,  China (No.20192BAB217003, No.20181ACG70022); 2019 Graduate Innovation Program of Nanchang Institute of Technology (No.YJSCX20190014)

摘要: 针对传统稀疏解混方法对丰度的稀疏性表征不充分及空间信息利用率低等问题,本文在分析迭代加权稀疏解混方法的基础上,提出了一种基于光谱加权协同稀疏和全变差正则化的高光谱解混方法.该方法一方面在协同稀疏解混的基础上引入光谱加权因子进一步刻画丰度系数的行稀疏性,以促进所有像元之间的联合稀疏性;另一方面引入各向异性全变差空间正则化促进图像同质区域的平滑性,以提高解混的准确性.通过交替方向乘子法求解该模型,通过迭代,利用内外部双循环迭代方法对光谱加权因子和丰度系数进行优化.模拟和真实的高光谱数据实验结果均表明本文提出的算法与现有同类算法相比能大幅提高混合像元分解的精度,在稀疏解混方面展现出了巨大的潜力.

关键词: 高光谱图像, 稀疏解混, 光谱加权协同稀疏, TV正则项, 空间信息

Abstract: In this work, we proposed a hyperspectral unmixing method based on the spectrally weighted collaborative sparsity and the total variation, aiming at alleviating the lack of the sparsity of abundance in traditional methods and fully exploiting the spatial information. On the one hand, the spectral factors are utilized to estimate the weights in order to enforce the sparsity of nonzero rows, thus improving the collaborative sparsity among all the pixels. On the other hand, the total variation based spatial regularization is employed to reinforce the smoothness within the homogenous regions, hence improving the accuracy of unmixing. The model is solved by the well-known alternating direction method of multiplier, in which the spectral factor based weights and the abundance coefficients are iteratively optimized using both the internal and external loops. The experimental results obtained from the simulated and the real datasets indicate that the proposed method could significantly improve the performance of unmixing compared to the other state-of-the-art methods.

Key words: hyperspectral imaging, sparse unmixing, spectral weighted collaborative sparse regression, total variation (TV), spatial information

中图分类号: