Robust unsupervised small area change detection from SAR imagery using deep learning

Zhang, Xinzheng; Su, Hang; Zhang, Ce; Gu, Xiaowei; Tan, Xiaoheng; Atkinson, Peter M.

doi:10.1016/j.isprsjprs.2021.01.004

Cited by 54 publications

(27 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this section, the IATO algorithm results based on the JR-KSVD algorithm are shown, and various techniques of change detection are compared to assess their benefits and drawbacks. A K-means-based change detection algorithm for SAR images (K-means) [39], [40], change detection method based on C-means clustering of trimmed fuzzy local information (RFLICM) [41], sparse autoencoder (SAE) and fuzzy c-means (FCM) clustering change detection algorithm (SAEFCM) [42], scale invariant feature transform (SIFT) feature point-based change detection algorithm (SIFT) [43], robust unsupervised small area change detection from SAR imagery using deep learning (RUSACD) [44], SAR image change detection via a Siamese adaptive fusion network (SAFNet) [45], automatic change detection in synthetic aperture radar images based on a PCA network (PCANet) (GaborP-CANet) [46], and change detection from SAR images based on neighborhood-based ratio and extreme learning machine (NRELM) [47] are compared with the methods proposed in this article.…”

Section: Table IV Change Detection Results Of Jr-ksvd and K-svdmentioning

confidence: 99%

SAR Image Change Detection Based on Joint Dictionary Learning With Iterative Adaptive Threshold Optimization

Zhang

et al. 2022

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

View full text Add to dashboard Cite

Synthetic aperture radar (SAR) image change detection is still a challenge due to inherent speckle noise and scarce datasets. This article proposes a joint-related dictionary learning algorithm based on the k-singular value decomposition (K-SVD) algorithm called JR-KSVD and an iterative adaptive threshold optimization (IATO) algorithm for unsupervised change detection. The JR-KSVD algorithm adds dictionary correlation learning to the K-SVD algorithm to generate a uniform initial dictionary for dual-temporal SAR images, thereby reducing the instability of sparse representations due to atomic correlations and enhancing the extraction of image edges and details. The IATO approach employs thresholds obtained by the "difference-log ratio" fusion image for indefinite residual energy minimization iterations to gradually shrink the threshold variation range and finally generate the change images, which have a high degree of adaptivity and strong real-time performance. Finally, experiments on six real datasets demonstrate that the proposed algorithm exhibits superior detection performance and robustness against some state-of-the-art algorithms.

show abstract

Section: Table IV Change Detection Results Of Jr-ksvd and K-svdmentioning

confidence: 99%

SAR Image Change Detection Based on Joint Dictionary Learning With Iterative Adaptive Threshold Optimization

Zhang

et al. 2022

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

View full text Add to dashboard Cite

show abstract

“…Next, into feature extraction process, the CWNN with DCGAN is used to map features to a low dimensional space for improving invariant property and decreasing dimension concurrently [18]. A deep CWNN is introduced for VPR and it is increased by a DCGAN to improve the sample size for the class.…”

Section: Design Of Cwnn-dcgan Modelmentioning

confidence: 99%

Seagull Optimization With Deep Learning Driven Condition Invariant Visual Place Recognition Model

Sasikumar¹,

Sathiamoorthy²

2022

INDJCSE

View full text Add to dashboard Cite

Visual place recognition (VPR) is most important topic in the computer vision (CV) and robotics community, whereas the task is for efficiently and accurately recognize the place of a provided query image. VPR still remains an open problem because of the several manners whereas the presence of real-world locations can change. As condition-invariant and viewpoint-invariant features were important features to long-term robust visual place-detection, the accomplishment of deep learning (DL) approaches from the CV domain triggered a kind of primary studies as its efficacy for VPR utilizing generic features in networks which are trained for another kind of detection tasks. In this aspect, this article develops a Seagull Optimization with Deep Learning Driven Condition Invariant Visual Place Recognition (SGODL-CIVPR) model. The presented SGODL-CIVPR model majorly involves the recognition of live places with respect to reference images. To accomplish this, the SGODL-CIVPR technique utilizes deep convolutional neural network based EfficientNet model to produce features, and the hyperparameter tuning process takes place using the SGO algorithm. Besides, convolutional wavelet neural network (CWNN) with deep convolutional generative adversarial network (DCGAN) is used to map features to low dimensional space for improving invariant properties and decrease dimensions concurrently. The performance validation of the SGODL-CIVPR model is tested using different datasets and the results demonstrate better performance over other methods.

show abstract

“…Currently, many networks dedicated to SAR images have achieved remarkable results in the change detection task 28 – 31 Moreover, there is a growing number of deep networks specifically designed for built-up area change detection 32 – 35 These approaches employ diverse strategies and have achieved limited success.…”

Section: Introductionmentioning

confidence: 99%

Segmentation-based VHR SAR images built-up area change detection: a coarse-to-fine approach

Zhu,

Wang,

You

2024

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

The change detection in built-up areas within very high resolution synthetic aperture radar images is a very challenging task due to speckle noise and geometric distortions caused by the unique imaging mechanism. To tackle this issue, we propose an object-based coarse-to-fine change detection method that integrates segmentation and uncertainty analysis techniques. First, we propose a multi-temporal joint multiscale segmentation method for generating multi-scale segmentation masks with hierarchical nested relationships. Second, we use the neighborhood ratio detector and Jensen-Shannon distance to produce both pixel-level and object-level change maps, respectively. These maps are fused using the Demeter-Shafer evidence theory, resulting in an initial change map. We then apply a threshold to classify parcels within the initial change map into three categories: changed, unchanged, and uncertain. Third, we perform uncertainty analysis and implement progressive classification by support vector machine for uncertain parcels, moving from coarse to fine segmentation levels. Finally, we integrate change maps across all scales to obtain the final change map. The proposed method is evaluated on three datasets from the GF-3 and ICEYE-X6 satellites. The results show that our approach outperforms alternative methods in extracting more comprehensive changed regions.

show abstract

Robust unsupervised small area change detection from SAR imagery using deep learning

Cited by 54 publications

References 41 publications

SAR Image Change Detection Based on Joint Dictionary Learning With Iterative Adaptive Threshold Optimization

SAR Image Change Detection Based on Joint Dictionary Learning With Iterative Adaptive Threshold Optimization

Seagull Optimization With Deep Learning Driven Condition Invariant Visual Place Recognition Model

Segmentation-based VHR SAR images built-up area change detection: a coarse-to-fine approach

Contact Info

Product

Resources

About