Unbiased Seamless SAR Image Change Detection Based on Normalized Compression Distance

Coca, Mihai; Anghel, Andrei; Datcu, Mihai

doi:10.1109/jstars.2019.2909143

Cited by 7 publications

(7 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Benefiting from the capability of all-weather and all-time Earth observation, the synthetic aperture radar (SAR) sensor has been used in numerous applications increasingly, including but not limited to urban planning, disaster monitoring, and land-cover/landuse (LCLU) analysis [1][2][3][4][5][6][7]. In reality, change detection (CD) in SAR images is crucial in these applications, which seeks to precisely identify the changed and unchanged parts by analyzing two or more SAR images acquired over the same geographic region at different times [2,3,[7][8][9]. However, SAR images exhibit diversified inherent characteristics, such as ubiquitous multiplicative speckle noise and geometrical distortions, that inevitably impose some challenges in SAR image CD [8][9][10].…”

Section: Introductionmentioning

confidence: 99%

“…In reality, change detection (CD) in SAR images is crucial in these applications, which seeks to precisely identify the changed and unchanged parts by analyzing two or more SAR images acquired over the same geographic region at different times [2,3,[7][8][9]. However, SAR images exhibit diversified inherent characteristics, such as ubiquitous multiplicative speckle noise and geometrical distortions, that inevitably impose some challenges in SAR image CD [8][9][10].…”

Section: Introductionmentioning

confidence: 99%

“…Secondly, most previous SAR image CD works [8,9,11,12,[14][15][16][17][18][19] focus on how to design an unsupervised model to achieve good performance, wherein clustering-based classifiers [9,15,17,18,28] are considered most representative and popular. Although their effectiveness and efficiency have been demonstrated, traditional clustering algorithms still suffer from some significant drawbacks, the major one of which is the poor ability in data fitting.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Hybrid Variability Aware Network (HVANet): A Self-Supervised Deep Framework for Label-Free SAR Image Change Detection

Wang

Liu

2022

Remote Sensing

View full text Add to dashboard Cite

Synthetic aperture radar (SAR) image change detection (CD) aims to automatically recognize changes over the same geographic region by comparing prechange and postchange SAR images. However, the detection performance is usually subject to several restrictions and problems, including the absence of labeled SAR samples, inherent multiplicative speckle noise, and class imbalance. More importantly, for bitemporal SAR images, changed regions tend to present highly variable sizes, irregular shapes, and different textures, typically referred to as hybrid variabilities, further bringing great difficulties to CD. In this paper, we argue that these internal hybrid variabilities can also be used for learning stronger feature representation, and we propose a hybrid variability aware network (HVANet) for completely unsupervised label-free SAR image CD by taking inspiration from recent developments in deep self-supervised learning. First, since different changed regions may exhibit hybrid variabilities, it is necessary to enrich distinguishable information within the input features. To this end, in shallow feature extraction, we generalize the traditional spatial patch (SP) feature to allow for each pixel in bitemporal images to be represented at diverse scales and resolutions, called extended SP (ESP). Second, with the carefully customized ESP features, HVANet performs local spatial structure information extraction and multiscale–multiresolution (MS-MR) information encoding simultaneously through a local spatial stream and a scale-resolution stream, respectively. Intrinsically, HVANet projects the ESP features into a new high-level feature space, where the change identification becomes easier. Third, to train the framework effectively, a self-supervision layer is attached to the top of the HVANet to enable the two-stream feature learning and recognition of changed pixels in the corresponding feature space, in a self-supervised manner. Experimental results on three low/medium-resolution SAR datasets demonstrate the effectiveness and superiority of the proposed framework in unsupervised SAR CD tasks.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Hybrid Variability Aware Network (HVANet): A Self-Supervised Deep Framework for Label-Free SAR Image Change Detection

Wang

Liu

2022

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…In [5], a brief SAR image change detection method based on NCD is projected. This method was further improved and applied for the detection of changes in TerraSAR-X images [6]. The distance matrix generated by applying NCD is used as input to supervised and unsupervised methods in order to obtain a change map in flooding scenarios.…”

Section: Introductionmentioning

confidence: 99%

GPU-Based Normalized Compression Distance for Satellite Images

Stoica¹,

Coca²

2022

J.Mil.Technol.

View full text Add to dashboard Cite

The measurement of the normalized compression distance (NCD) between two objects is computed by counting the number of common patterns that appear in the representation of compressed objects. Classical Lempel-Ziv (LZ) algorithms do not achieve high compression speeds on images because they require sequentially traversing the linearized version of object to build the corresponding dictionary, i.e., dynamic approach to optimize the encoding based upon the particular input. This paper proposes a faster approach for computing the NCD metric in a parallel manner by estimating compressions on GPU and multi-thread schemes. The goal of calculating distances is to use them in clustering and classification of input images. The tests are run on the UC Merced Land Use and BigEarthNet datasets, which are both well-known and frequently used in the field of remote sensing.

show abstract

“…The combined use of Sentinel's radar and optical data allows for land monitoring temporally, because Sentinel's frequent revisit time makes it possible to produce maps of land change during short time intervals [6,7]. Classification of land change requires procedures depending on the data source, such as hyperspectral images [8,9] or Synthetic Aperture Radar data [10,11]. Researchers can apply ready-to-use LULC datasets, like the Corine Land Cover database available in European countries.…”

Section: Introductionmentioning

confidence: 99%

Effects of Category Aggregation on Land Change Simulation Based on Corine Land Cover Data

et al. 2020

View full text Add to dashboard Cite

Several factors influence the performance of land change simulation models. One potentially important factor is land category aggregation, which reduces the number of categories while having the potential to reduce also the size of apparent land change in the data. Our article compares how four methods to aggregate Corine Land Cover categories influence the size of land changes in various spatial extents and consequently influence the performance of 114 Cellular Automata-Markov simulation model runs. We calculated the reference change during the calibration interval, the reference change during the validation interval and the simulation change during the validation interval, along with five metrics of simulation performance, Figure of Merit and its four components: Misses, Hits, Wrong Hits and False Alarms. The Corine Standard Level 1 category aggregation reduced change more than any of the other aggregation methods. The model runs that used the Corine Standard Level 1 aggregation method tended to return lower sizes of changing areas and lower values of Misses, Hits, Wrong Hits and False Alarms, where Hits are correctly simulated changes. The behavior-based aggregation method maintained the most change while using fewer categories compared to the other aggregation methods. We recommend an aggregation method that maintains the size of the reference change during the calibration and validation intervals while reducing the number of categories, so the model uses the largest size of change while using fewer than the original number of categories.

show abstract

Unbiased Seamless SAR Image Change Detection Based on Normalized Compression Distance

Cited by 7 publications

References 19 publications

Hybrid Variability Aware Network (HVANet): A Self-Supervised Deep Framework for Label-Free SAR Image Change Detection

Hybrid Variability Aware Network (HVANet): A Self-Supervised Deep Framework for Label-Free SAR Image Change Detection

GPU-Based Normalized Compression Distance for Satellite Images

Effects of Category Aggregation on Land Change Simulation Based on Corine Land Cover Data

Contact Info

Product

Resources

About