Tensor Decomposition-Inspired Convolutional Autoencoders for Hyperspectral Anomaly Detection
文献类型:期刊论文
作者 | Sun, Bangyong4; Zhao, Zhe4; Liu, Di3; Gao, Xiaomei2; Yu, Tao1 |
刊名 | IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING |
出版日期 | 2022-06-21 |
卷号 | 15页码:4990-5000 |
ISSN号 | 1939-1404;2151-1535 |
关键词 | Anomaly detection Hyperspectral imaging Tensors Detectors Feature extraction Training Neural networks Anomaly detection hyperspectral image (HSI) tensor decomposition network |
DOI | 10.1109/JSTARS.2022.3184789 |
产权排序 | 4 |
英文摘要 | Anomaly detection from hyperspectral images (HSI) is an important task in the remote sensing domain. Considering the three-order characteristics of HSI, many tensor decomposition based hyperspectral anomaly detection (HAD) models have been proposed and drawn much attention during the past decades. However, as most tensor decomposition based detectors are directly performed on the original HSI, the detection accuracy is usually limited due to the high-dimension and noise corruption of the HSI. Benefiting from the good capacity of autoencoders (AE) for feature extraction, in this article, an enhanced tensor decomposition-inspired convolutional AE for HAD is proposed to address those problems, named TDNet. Within the proposed TDNet, the traditional canonical-polyadic (CP) tensor decomposition model is innovatively alternated by a deep neural network (DNN), and the DNN tensor decomposition model performs more stably and robustly for noise. Specifically, a potential abnormal pixels remove strategy is first built to obtain the background training sets. Then, a DNN tensor decomposition-inspired convolutional AE is used to recover the original background information, which consists of an encoder, a low-rank tensor decomposition network, and a decoder. Finally, the residual errors between input HSI and recovered background are used for anomaly detection. Extensive experiments demonstrate the superiority of the TDNet in terms of both AUC values and ROC curves. |
语种 | 英语 |
出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
WOS记录号 | WOS:000820509500001 |
源URL | [http://ir.opt.ac.cn/handle/181661/96050] |
专题 | 西安光学精密机械研究所_光学影像学习与分析中心 |
通讯作者 | Sun, Bangyong |
作者单位 | 1.Chinese Acad Sci, Xian Inst Opt & Precis Mech, Key Lab Spectral Imaging Technol CAS, Xian 710119, Peoples R China 2.China Natl Adm Coal Geol, Xian Mapping & Printing, Xian 710199, Peoples R China 3.Beijing Technol & Business Univ, Beijing Key Lab Big Data Technol Food Safety, Beijing 100048, Peoples R China 4.Xian Univ Technol, Sch Printing Packaging & Digital Media, Xian 710048, Peoples R China |
推荐引用方式 GB/T 7714 | Sun, Bangyong,Zhao, Zhe,Liu, Di,et al. Tensor Decomposition-Inspired Convolutional Autoencoders for Hyperspectral Anomaly Detection[J]. IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING,2022,15:4990-5000. |
APA | Sun, Bangyong,Zhao, Zhe,Liu, Di,Gao, Xiaomei,&Yu, Tao.(2022).Tensor Decomposition-Inspired Convolutional Autoencoders for Hyperspectral Anomaly Detection.IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING,15,4990-5000. |
MLA | Sun, Bangyong,et al."Tensor Decomposition-Inspired Convolutional Autoencoders for Hyperspectral Anomaly Detection".IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 15(2022):4990-5000. |
入库方式: OAI收割
来源:西安光学精密机械研究所
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。