High-Resolution Remote Sensing Bitemporal Image Change Detection Based on Feature Interaction and Multitask Learning

Zhao, Chunhui; Tang, Yingjie; Feng, Shou; Fan, Yuanze; Li, Wei; Tao, Ran; Zhang, Lifu

doi:10.1109/tgrs.2023.3275140

Cited by 14 publications

(3 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…TINYCD directly shares the weight of the existing network to achieve the purpose of the Siamese network [34]. The method based on feature interaction and multitask learning (FMCD) [35] can improve the ability to detect changes in complex scenes, by modeling the context information of features through a multilevel feature interaction module, so as to obtain representative features, and to improve the sensitivity of the model to changes. The change gradient image (CGI) [36] first embeds a multiscale information attentional module in U-Net to achieve multiscale information and adds the position channel attention module to pay more attention to the spectral and spatial information in the multiscale fused feature map.…”

Section: Deep-learning Methods For Change Detection (Cd)mentioning

confidence: 99%

HSAA-CD: A Hierarchical Semantic Aggregation Mechanism and Attention Module for Non-Agricultural Change Detection in Cultivated Land

Li,

Zhou,

Zhang

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

Cultivated land plays a fundamental role in the sustainable development of the world. Monitoring the non-agricultural changes is important for the development of land-use policies. A bitemporal image transformer (BIT) can achieve high accuracy for change detection (CD) tasks and also become a key scientific tool to support decision-making. Because of the diversity of high-resolution RSIs in series, the complexity of agricultural types, and the irregularity of hierarchical semantics in different types of changes, the accuracy of non-agricultural CD is far below the need for the management of the land and for resource planning. In this paper, we proposed a novel non-agricultural CD method to improve the accuracy of machine processing. First, multi-resource surveying data are collected to produce a well-tagged dataset with cultivated land and non-agricultural changes. Secondly, a hierarchical semantic aggregation mechanism and attention module (HSAA) bitemporal image transformer method named HSAA-CD is performed for non-agricultural CD in cultivated land. The proposed HSAA-CD added a hierarchical semantic aggregation mechanism for clustering the input data for U-Net as the backbone network and an attention module to improve the feature edge. Experiments were performed on the open-source LEVIR-CD and WHU Building-CD datasets as well as on the self-built RSI dataset. The F1-score, intersection over union (IoU), and overall accuracy (OA) of these three datasets were 88.56%, 84.29%, and 68.50%; 79.84%, 73.41%, and 59.29%; and 98.83%, 98.39%, and 93.56%, respectively. The results indicated that the proposed HSAA-CD method outperformed the BIT and some other state-of-the-art methods and proved to be suitable accuracy for non-agricultural CD in cultivated land.

show abstract

Section: Deep-learning Methods For Change Detection (Cd)mentioning

confidence: 99%

HSAA-CD: A Hierarchical Semantic Aggregation Mechanism and Attention Module for Non-Agricultural Change Detection in Cultivated Land

Li,

Zhou,

Zhang

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…At different scales global information was obtained using the U-Net. SNUNet was effective but the error was significantly increased in the complex ground object information [46]. This method was tested using the DSIFN-CD, SYSUCD, and WHU-CD datasets.…”

Section: U-netmentioning

confidence: 99%

Change detection using multispectral satellite images: a systematic review of literature

Vasantrao,

Gupta,

Mishra

et al. 2024

Bulletin EEI

View full text Add to dashboard Cite

Change detection (CD) provides information about the changes on earth’s surface over a period of time. Many algorithms have been proposed over the years for effective CD of satellite imagery. This paper presents the steps to preprocess the captured satellite images, which can be used to perform predictive analysis of earth’s surface by CD techniques. To design a highly effective system for CD, it is recommended that algorithm designers select efficient algorithms from any given application. Thus, this paper presents and investigates the review of most appropriate literature on CD, where CD techniques have been presented into three groups; i) thresholding, ii) clustering, and iii) deep learning. The first two categories mainly rely on the traditional machine learning, whereas the last one exploits the state-of-the-art deep learning models. At the end, the standard methods are summarized based on advantages, limitation, and research gap.

show abstract

“…In the domain of semantic change detection, multi-task structures have gained prominence and are presently considered the prevailing approach [24]. For instance, in a study by Zhao, et al [25], a change detection model coupled with domain adaption as an auxiliary task was proposed to mitigate the impact of domain shifts and irrelevant changes. Similarly, in a study focused on building change detection, the change detection loss function is constrained by auxiliary building detection tasks, thereby leveraging intrinsically correlated features within building footprints for the detection of fine-grained building changes [26].…”

Section: A Semantic Change Detectionmentioning

confidence: 99%

Adaptive Self-Paced Collaborative and 3-D Adversarial Multitask Network for Semantic Change Detection Using Zhuhai-1 Orbita Hyperspectral Remote Sensing Imagery

Wen,

Huang,

Yang

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

In recent years, numerous change detection methodologies have been proposed, with a predominant focus on binary change detection. Furthermore, there exists a paucity of research addressing semantic change detection in scenarios where solely binary change labels are available. This article introduces a multi-task network for semantic change detection. First, 3D ResUnet model is employed to generate initial multitemporal land cover results through post classification comparison (PCC). Subsequently, the multi-task network, encompassing two sub-tasks-binary change detection and multi-temporal semantic segmentation-is proposed. Specifically, the shared branch of the network employs 3D residual blocks to extract joint spectral-spatial features. In the subsequent taskspecific branch, a 3D GAN is incorporated for the binary change detection task to enhance the discrimination ability of latent high-level features for changes. Novel adaptive self-paced learning and certainty-weighted focal loss are proposed for multitemporal semantic segmentation to mitigate adverse effects from noisy semantic labels by considering sample complexity and reliability in the network optimization process. Experiments conducted on the Orbita Hyperspectral Dataset in the Xiong'an New Area demonstrate the superior performance of the proposed method, achieving 99.28% and 76.60% for Overall Accuracy (OA) and Kappa, respectively. This outperformance is notable when compared to other methods, such as Str4 and Bi-SRNet, showing an increase of 39.82% and 54.17% for Kappa. Moreover, comparative experiments on SECOND further confirm the advantage of the proposed method, achieving 54.62% for Kappa and outperforming other comparative methos like Bi-SRNet

show abstract

High-Resolution Remote Sensing Bitemporal Image Change Detection Based on Feature Interaction and Multitask Learning

Cited by 14 publications

References 34 publications

HSAA-CD: A Hierarchical Semantic Aggregation Mechanism and Attention Module for Non-Agricultural Change Detection in Cultivated Land

HSAA-CD: A Hierarchical Semantic Aggregation Mechanism and Attention Module for Non-Agricultural Change Detection in Cultivated Land

Change detection using multispectral satellite images: a systematic review of literature

Adaptive Self-Paced Collaborative and 3-D Adversarial Multitask Network for Semantic Change Detection Using Zhuhai-1 Orbita Hyperspectral Remote Sensing Imagery

Contact Info

Product

Resources

About