remotesensing-logo

Journal Browser

Journal Browser

Target Recognition and Change Detection for High-Resolution Remote Sensing Images

A special issue of Remote Sensing (ISSN 2072-4292). This special issue belongs to the section "Remote Sensing Image Processing".

Deadline for manuscript submissions: 15 October 2024 | Viewed by 4233

Special Issue Editors


E-Mail Website
Guest Editor
State Key Laboratory of Information Engineering in Surveying, Mapping, and Remote Sensing, Wuhan University, Wuhan, China
Interests: multi-temporal image processing and change detection; hyper-spectral image processing; high-resolution image understanding; satellite video tracking; multi-source data fusion; machine learning; computer vision; urban remote sensing

E-Mail Website
Guest Editor
School of Information Engineering, Ningxia University, Yinchuan 750021, China
Interests: high spatial resolution remote sensing image classification; change detection; hyperspectral remote sensing image interpretation; machine learning

E-Mail Website
Guest Editor
School of Electronic Information, Wuhan University, Wuhan 430072, China
Interests: objection detection and recognition; multimodal image registration; cross-modal geo-localization
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

The fast development of remote sensing platforms brings further improvement in the resolution of remote sensing images. High-resolution remote sensing images contain more detailed spatial, spectral, and temporal information of ground landscapes. Recognizing the targets and the changes from multi-source high-resolution remote sensing data has becomes an important topic for Earth observation techniques in many applications. However, this task still encounters several challenges: (1) From the aspect of the images: though high-resolution images make it possible to monitor more types of targets and changes with precise boundaries, the problem of spatial and spectral variability caused by different acquiring conditions (illumination, angle, atmosphere) and sensor characteristics is also severe. (2) From the aspect of the landscapes: the targets on the ground inherently show more complex shapes and higher diversities, which makes it difficult to model their information by traditional methods. (3) From the aspect of applications: manual labelling for training and verifying with high time- and labour-costs, has limited the universality of applying high-resolution remote sensing data to multiple fields. Therefore, it is a very interesting and crucial topic.

This Special Issue aims to focus on discussing the theoretical frontiers and technical problems in target recognition and change detection for high-resolution remote sensing images and provide a platform for researchers to show their recent contributions.

  • Target recognition and change detection with high spatial, spectral, and temporal resolution remote sensing images.
  • Advanced interpretation of high-resolution images by unsupervised, semi-supervised, weakly supervised, and self-supervised mechanisms.
  • Datasets and benchmarks for target recognition and change detection with high-resolution remote sensing images.
  • Advances in machine learning and deep learning techniques for high-resolution remote sensing image processing.
  • The applications of high-resolution remote sensing data in various fields. 

Prof. Dr. Chen Wu
Dr. Pengyuan Lv
Dr. Naoto Yokoya
Prof. Dr. Wen Yang
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Remote Sensing is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2700 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • target recognition
  • change detection
  • high-resolution remote sensing
  • deep learning
  • machine learning

Published Papers (6 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Research

21 pages, 6147 KiB  
Article
SDRnet: A Deep Fusion Network for ISAR Ship Target Recognition Based on Feature Separation and Weighted Decision
by Jie Deng and Fulin Su
Remote Sens. 2024, 16(11), 1920; https://0-doi-org.brum.beds.ac.uk/10.3390/rs16111920 - 27 May 2024
Viewed by 414
Abstract
Existing methods for inverse synthetic aperture radar (ISAR) target recognition typically rely on a single high-resolution radar signal type, such as ISAR images or high-resolution range profiles (HRRPs). However, ISAR images and HRRP data offer representations of targets across different aspects, each containing [...] Read more.
Existing methods for inverse synthetic aperture radar (ISAR) target recognition typically rely on a single high-resolution radar signal type, such as ISAR images or high-resolution range profiles (HRRPs). However, ISAR images and HRRP data offer representations of targets across different aspects, each containing valuable information crucial for radar target recognition. Moreover, the process of generating ISAR images inherently facilitates the acquisition of HRRP data, ensuring timely data collection. Therefore, to fully leverage the different information from both HRRP data and ISAR images and enhance ISAR ship target recognition performance, we propose a novel deep fusion network named the Separation-Decision Recognition network (SDRnet). First, our approach employs a convolutional neural network (CNN) to extract initial feature vectors from ISAR images and HRRP data. Subsequently, a feature separation module is employed to derive a more robust target representation. Finally, we introduce a weighted decision module to enhance overall predictive performance. We validate our method using simulated and measured data containing ten categories of ship targets. The experimental results confirm the effectiveness of our approach in improving ISAR ship target recognition. Full article
Show Figures

Figure 1

32 pages, 10464 KiB  
Article
The Cost of Urban Renewal: Annual Construction Waste Estimation via Multi-Scale Target Information Extraction and Attention-Enhanced Networks in Changping District, Beijing
by Lei Huang, Shaofu Lin, Xiliang Liu, Shaohua Wang, Guihong Chen, Qiang Mei and Zhe Fu
Remote Sens. 2024, 16(11), 1889; https://0-doi-org.brum.beds.ac.uk/10.3390/rs16111889 - 24 May 2024
Viewed by 370
Abstract
Construction waste is an inevitable byproduct of urban renewal, causing severe pressure on the environment, health, and ecology. Accurately estimating the production of construction waste is crucial for assessing the consumption of urban renewal. However, traditional manual estimation methods rely heavily on statistical [...] Read more.
Construction waste is an inevitable byproduct of urban renewal, causing severe pressure on the environment, health, and ecology. Accurately estimating the production of construction waste is crucial for assessing the consumption of urban renewal. However, traditional manual estimation methods rely heavily on statistical data and historical experience, which lack flexibility in practical applications and are time-consuming and labor-intensive. In addition, their accuracy and timeliness need to be improved urgently. Fortunately, with the advantages of high-resolution remote sensing images (HRSIs) such as strong timeliness, large amounts of information, and macroscopic observations, they are suitable for the large-scale dynamic change detection of construction waste. However, the existing deep learning models have a relatively poor ability to extract and fuse features for small and multi-scale targets, and it is difficult to deal with irregularly shaped and fragmented detection areas. Therefore, this study proposes a Multi-scale Target Attention-Enhanced Network (MT-AENet), which is used to dynamically track and detect changes in buildings and construction waste disposal sites through HRSIs and accurately estimate the annual production of urban construction waste. The MT-AENet introduces a novel encoder–decoder architecture. In the encoder, ResNet-101 is utilized to extract high-level semantic features. A depthwise separable-atrous spatial pyramid pooling (DS-ASPP) module with different dilation rates is constructed to address insufficient receptive fields, resolving the issue of discontinuous holes when extracting large targets. A dual-attention mechanism module (DAMM) is employed to better preserve positional and channel details. In the decoder, multi-scale feature fusion (MS-FF) is utilized to capture contextual information, integrating shallow and intermediate features of the backbone network, thereby enhancing extraction capabilities in complex scenes. The MT-AENet is used to extract buildings and construction waste at different periods in the study area, and the actual production and landfill volume of construction waste are calculated based on area changes, indirectly measuring the rate of urban construction waste resource conversion. The experimental results in Changping District, Beijing demonstrate that the MT-AENet outperforms existing baseline networks in extracting buildings and construction waste. The results of this study are validated according to government statistical standards, providing a promising direction for efficiently analyzing the consumption of urban renewal. Full article
Show Figures

Graphical abstract

27 pages, 7580 KiB  
Article
Change Detection Based on Existing Vector Polygons and Up-to-Date Images Using an Attention-Based Multi-Scale ConvTransformer Network
by Shengli Wang, Yihu Zhu, Nanshan Zheng, Wei Liu, Hua Zhang, Xu Zhao and Yongkun Liu
Remote Sens. 2024, 16(10), 1736; https://0-doi-org.brum.beds.ac.uk/10.3390/rs16101736 - 14 May 2024
Viewed by 671
Abstract
Vector polygons represent crucial survey data, serving as a cornerstone of national geographic censuses and forming essential data sources for detecting geographical changes. The timely update of these polygons is vital for governmental decision making and various industrial applications. However, the manual intervention [...] Read more.
Vector polygons represent crucial survey data, serving as a cornerstone of national geographic censuses and forming essential data sources for detecting geographical changes. The timely update of these polygons is vital for governmental decision making and various industrial applications. However, the manual intervention required to update existing vector polygons using up-to-date high-resolution remote sensing (RS) images poses significant challenges and incurs substantial costs. To address this, we propose a novel change detection (CD) method for land cover vector polygons leveraging high-resolution RS images and deep learning techniques. Our approach begins by employing the boundary-preserved masking Simple Linear Iterative Clustering (SLIC) algorithm to segment RS images. Subsequently, an adaptive cropping approach automatically generates an initial sample set, followed by denoising using the efficient Visual Transformer and Class-Constrained Density Peak-Based (EViTCC-DP) method, resulting in a refined training set. Finally, an enhanced attention-based multi-scale ConvTransformer network (AMCT-Net) conducts fine-grained scene classification, integrating change rules and post-processing methods to identify changed vector polygons. Notably, our method stands out by employing an unsupervised approach to denoise the sample set, effectively transforming noisy samples into representative ones without requiring manual labeling, thus ensuring high automation. Experimental results on real datasets demonstrate significant improvements in model accuracy, with accuracy and recall rates reaching 92.08% and 91.34%, respectively, for the Nantong dataset, and 93.51% and 92.92%, respectively, for the Guantan dataset. Moreover, our approach shows great potential in updating existing vector data while effectively mitigating the high costs associated with acquiring training samples. Full article
Show Figures

Figure 1

17 pages, 7615 KiB  
Article
Instantaneous Frequency Extraction for Nonstationary Signals via a Squeezing Operator with a Fixed-Point Iteration Method
by Zhen Li, Zhaoqi Gao, Fengyuan Sun, Jinghuai Gao and Wei Zhang
Remote Sens. 2024, 16(8), 1412; https://0-doi-org.brum.beds.ac.uk/10.3390/rs16081412 - 16 Apr 2024
Viewed by 512
Abstract
The instantaneous frequency (IF) is an important feature for the analysis of nonstationary signals. For IF estimation, the time–frequency representation (TFR)-based algorithm is used in a common class of methods. TFR-based methods always need the representation concentrated around the “true” IFs and the [...] Read more.
The instantaneous frequency (IF) is an important feature for the analysis of nonstationary signals. For IF estimation, the time–frequency representation (TFR)-based algorithm is used in a common class of methods. TFR-based methods always need the representation concentrated around the “true” IFs and the number of components within the signal. In this paper, we propose a novel method to adaptively estimate the IFs of nonstationary signals, even for weak components of the signals. The proposed technique is not based on the TFR: it is based on the frequency estimation operator (FEO), and the short-time Fourier transform (STFT) is used as its basis. As we know, the FRO is an exact estimation of the IF for weak frequency-modulated (FM) signals, but is not appropriate for strong FM modes. Through theoretical derivation, we determine that the fixed points of the FEOwith respect to the frequency are equivalent to the ridge of the STFT spectrum. Furthermore, the IF of the linear chirp signals is just the fixed points of the FEO. Therefore, we apply the fixed-point algorithm to the FEO to realize the precise and reliable estimation of the IF, even for highly FM signals. Finally, the results using synthetic and real signals show the utility of the proposed method for IF estimation and that it is more robust than the compared method. It should be noted that the proposed method employing the FEO only computes the first-order differential of the STFT for the chirp-like signals, while it can provide a result derived using the second-order estimation operator. Moreover, this new method is effective for the IF estimation of weak components within a signal. Full article
Show Figures

Figure 1

23 pages, 2284 KiB  
Article
MDANet: A High-Resolution City Change Detection Network Based on Difference and Attention Mechanisms under Multi-Scale Feature Fusion
by Shanshan Jiang, Haifeng Lin, Hongjin Ren, Ziwei Hu, Liguo Weng and Min Xia
Remote Sens. 2024, 16(8), 1387; https://0-doi-org.brum.beds.ac.uk/10.3390/rs16081387 - 14 Apr 2024
Cited by 1 | Viewed by 651
Abstract
In the domains of geographic information systems and remote sensing image analysis, change detection is vital for examining surface variations in high-resolution remote sensing pictures. However, the intricate texture characteristics and rich details found in high-resolution remote sensing photos are difficult for conventional [...] Read more.
In the domains of geographic information systems and remote sensing image analysis, change detection is vital for examining surface variations in high-resolution remote sensing pictures. However, the intricate texture characteristics and rich details found in high-resolution remote sensing photos are difficult for conventional change detection systems to deal with. Target misdetection, missed detections, and edge blurring are further problems with current deep learning-based methods. This research proposes a high-resolution city change detection network based on difference and attention mechanisms under multi-scale feature fusion (MDANet) to address these issues and improve the accuracy of change detection. First, to extract features from dual-temporal remote sensing pictures, we use the Siamese architecture as the encoder network. The Difference Feature Module (DFM) is employed to learn the difference information between the dual-temporal remote sensing images. Second, the extracted difference features are optimized with the Attention Refinement Module (ARM). The Cross-Scale Fusion Module (CSFM) combines and enhances the optimized attention features, effectively capturing subtle differences in remote sensing images and learning the finer details of change targets. Finally, thorough tests on the BTCDD dataset, LEVIR-CD dataset, and CDD dataset show that the MDANet algorithm performs at a cutting-edge level. Full article
Show Figures

Figure 1

18 pages, 1580 KiB  
Article
SFDA-CD: A Source-Free Unsupervised Domain Adaptation for VHR Image Change Detection
by Jingxuan Wang and Chen Wu
Remote Sens. 2024, 16(7), 1274; https://0-doi-org.brum.beds.ac.uk/10.3390/rs16071274 - 4 Apr 2024
Viewed by 660
Abstract
Deep models may have disappointing performance in real applications due to the domain shifts in data distributions between the source and target domain. Although a few unsupervised domain adaptation methods have been proposed to make the pre-train models effective on target domain datasets, [...] Read more.
Deep models may have disappointing performance in real applications due to the domain shifts in data distributions between the source and target domain. Although a few unsupervised domain adaptation methods have been proposed to make the pre-train models effective on target domain datasets, constraints like data privacy, security, and transmission limits restrict access to VHR remote sensing images, making existing unsupervised domain adaptation methods almost ineffective in specific change detection areas. Therefore, we propose a source-free unsupervised domain adaptation change detection structure to complete specific change detection tasks, using only the pre-trained source model and unlabelled target data. The GAN-based source generation component is designed to generate synthetic source data, which, to some extent, reflects the distribution of the source domain. Moreover, these data can be utilised in model knowledge transfer. The model adaptation component facilitates knowledge transfer between models by minimising the differences between deep features, using AAM (Attention Adaptation Module) to extract the difference between high-level features, meanwhile we proposed ISM (Intra-domain Self-supervised Module) to train target model in a self-supervised strategy in order to improve the knowledge adaptation. Our SFDA-CD framework demonstrates superior accuracy over existing unsupervised domain adaptation change detection methods, which has 0.6% cIoU and 1.5% F1 score up in cross-regional tasks and 1.4% cIoU and 1.9% F1 score up in cross-scenario tasks, proving that it can effectively reduce the domain shift between the source and target domains even without access to source data. Additionally, it can facilitate knowledge transfer from the source model to the target model. Full article
Show Figures

Figure 1

Back to TopTop