• Korean Journal of Remote Sensing
• /
• v.29 no.5
• /
• pp.569-579
• /
• 2013

In this study, land-cover classification was implemented using features extracted from multi-temporal SAR data through 3D wavelet transform and the applicability of the 3D wavelet transform as a feature extraction approach was evaluated. The feature extraction stage based on 3D wavelet transform was first carried out before the classification and the extracted features were used as input for land-cover classification. For a comparison purpose, original image data without the feature extraction stage and Principal Component Analysis (PCA) based features were also classified. Multi-temporal Radarsat-1 data acquired at Dangjin, Korea was used for this experiment and five land-cover classes including paddy fields, dry fields, forest, water, and built up areas were considered for classification. According to the discrimination capability analysis, the characteristics of dry field and forest were similar, so it was very difficult to distinguish these two classes. When using wavelet-based features, classification accuracy was generally improved except built-up class. Especially the improvement of accuracy for dry field and forest classes was achieved. This improvement may be attributed to the wavelet transform procedure decomposing multi-temporal data not only temporally but also spatially. This experiment result shows that 3D wavelet transform would be an effective tool for feature extraction from multi-temporal data although this procedure should be tested to other sensors or other areas through extensive experiments.

• Korean Journal of Remote Sensing
• /
• v.27 no.4
• /
• pp.435-444
• /
• 2011

The land cover map derived from spectral features of high resolution optical images has low spectral resolution and heterogeneity in the same land cover class. For this reason, despite the same land cover class, the land cover can be classified into various land cover classes especially in vegetation area. In order to overcome these problems, detailed vegetation classification is applied to optical satellite image and SAR(Synthetic Aperture Radar) integrated data in vegetation area which is the result of pre-classification from optical image. The pre-classification and vegetation classification were performed with MLC(Maximum Likelihood Classification) method. The hierarchical land cover classification was proposed from fusion of detailed vegetation classes and non-vegetation classes of pre-classification. We can verify the facts that the proposed method has higher accuracy than not only general SAR data and GLCM(Gray Level Co-occurrence Matrix) texture integrated methods but also hierarchical GLCM integrated method. Especially the proposed method has high accuracy with respect to both vegetation and non-vegetation classification.

• Korean Journal of Geomatics
• /
• v.2 no.1
• /
• pp.37-46
• /
• 2002

Accurate classification of water area is an preliminary step to accurately analyze the flooded area and damages caused by flood. This step is especially useful for monitoring the region where annually repeating flood is a problem. The accurate estimation of flooded area can ultimately be utilized as a primary source of information for the policy decision. Although SAR (Synthetic Aperture Radar) imagery with its own energy source is sensitive to the water area, its shadow effect similar to the reflectance signature of the water area should be carefully checked before accurate classification. Especially when we want to identify small flood area with mountainous environment, the step for removing shadow effect turns out to be essential in order to accurately classify the water area from the SAR imagery. In this paper, the flood area was classified and monitored using multi-temporal RADARSAT SAR images of Ok-Chun and Bo-Eun located in Chung-Book Province taken in 12th (during the flood) and 19th (after the flood) of August, 1998. We applied several steps of geometric and radiometric calculations to the SAR imagery. First we reduced the speckle noise of two SAR images and then calculated the radar backscattering coefficient $(\sigma^0)$. After that we performed the ortho-rectification via satellite orbit modeling developed in this study using the ephemeris information of the satellite images and ground control points. We also corrected radiometric distortion caused by the terrain relief. Finally, the water area was identified from two images and the flood area is calculated accordingly. The identified flood area is analyzed by overlapping with the existing land use map.

• Korean Journal of Remote Sensing
• /
• v.35 no.6_1
• /
• pp.999-1009
• /
• 2019

The importance of high-resolution sea ice maps of the Arctic Ocean is increasing due to the possibility of pioneering North Pole Routes and the necessity of precise climate prediction models. In this study,sea ice classification algorithms for two deep learning models were examined using Sentinel-1 A/B SAR data to generate high-resolution sea ice classification maps. Based on current ice charts, three classes (Open Water, First Year Ice, Multi Year Ice) of training data sets were generated by Arctic sea ice and remote sensing experts. Ten sea ice classification algorithms were generated by combing two deep learning models (i.e. Simple CNN and Resnet50) and five cases of input bands including incident angles and thermal noise corrected HV bands. For the ten algorithms, analyses were performed by comparing classification results with ground truth points. A confusion matrix and Cohen's kappa coefficient were produced for the case that showed best result. Furthermore, the classification result with the Maximum Likelihood Classifier that has been traditionally employed to classify sea ice. In conclusion, the Convolutional Neural Network case, which has two convolution layers and two max pooling layers, with HV and incident angle input bands shows classification accuracy of 96.66%, and Cohen's kappa coefficient of 0.9499. All deep learning cases shows better classification accuracy than the classification result of the Maximum Likelihood Classifier.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.15 no.2
• /
• pp.103-112
• /
• 2012

We demonstrate the preliminary results of ship detection application using synthetic aperture radar (SAR) and automatic identification system (AIS) together. Multi-frequency and multi-temporal SAR images such as TerraSAR-X and Cosmo-SkyMed (X-band), and Radarsat-2 (C-band) are acquired over the West Sea in South Korea. In order to compare with SAR data, we also collected an AIS data. The SAR data are pre-processed considering by the characteristics of scattering mechanism as for sea surface. We proposed the "Adaptive Threshold Algorithm" for classification ship efficiently. The analyses using the combination of the SAR and AIS data with time series will be very useful to ship detection or tracing of the ship.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.30 no.4
• /
• pp.275-281
• /
• 2019

A synthetic aperture radar(SAR) satellite is more effective than an optical satellite for target analysis because an SAR satellite can provide two-dimensional electromagnetic scattering distribution of a target during all-weather and day-and-night operations. To conduct target analysis while considering the earth observation interval of an SAR satellite, observing a specific area as wide as possible would be advantageous. However, wider the observation area, worse is the resolution of the associated SAR satellite image. Although conventional methods for improving the resolution of radar images can be employed for addressing this issue, few studies have been conducted for improving the resolution of SAR satellite images and analyzing the performance. Hence, in this study, the applicability of conventional methods to SAR satellite images is investigated. SAR target detection was first applied to Korea Multipurpose Satellite-5(KOMPSAT-5) SAR images provided by Korea Aerospace Research Institute for extracting target responses. Extrapolation, RELAX, and MUSIC algorithms were subsequently applied to the target responses for improving the resolution, and the corresponding performance was thereby analyzed.

• Korean Journal of Remote Sensing
• /
• v.36 no.6_1
• /
• pp.1303-1322
• /
• 2020

Indonesia is more prone to natural disasters due to its geological condition under the three main plates, making Indonesia experience frequent seismic activity, causing earthquakes, volcanic eruption, and tsunami. Those disasters could lead to other disasters such as landslides, floods, land subsidence, and coastal inundation. Monitoring those disasters could be essential to predict and prevent damage to the environment. We reviewed the application of remote sensing and Geographic Information System (GIS) for detecting natural disasters in the case of Indonesia, based on 43 articles. The remote sensing and GIS method will be focused on InSAR techniques, image classification, and susceptibility mapping. InSAR method has been used to monitor natural disasters affecting the deformation of the earth's surface in Indonesia, such as earthquakes, volcanic activity, and land subsidence. Monitoring landslides in Indonesia using InSAR techniques has not been found in many studies; hence it is crucial to monitor the unstable slope that leads to a landslide. Image classification techniques have been used to monitor pre-and post-natural disasters in Indonesia, such as earthquakes, tsunami, forest fires, and volcano eruptions. It has a lack of studies about the classification of flood damage in Indonesia. However, flood mapping was found in susceptibility maps, as many studies about the landslide susceptibility map in Indonesia have been conducted. However, a land subsidence susceptibility map was the one subject to be studied more to decrease land subsidence damage, considering many reported cases found about land subsidence frequently occur in several cities in Indonesia.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.3
• /
• pp.212-218
• /
• 2013

This paper presents an SVD(Singular Value Decomposition)-Pseudo Spectrum method for SAR (Synthetic Aperture Radar) imaging. The purpose of this work is to improve resolution and target separability of SAR images. This paper proposes SVD-Pseudo Spectrum method whose advantages are noise robustness, reduction of sidelobes and high resolution of spectral estimation. SVD-Pseudo Spectrum method uses Hankel Matrix of signal components and SVD (Singular Value Decomposition) method. In this paper, it is demonstrated that the SVD-Pseudo Spectrum method shows better performance than the matched filtering method and the conventional super-resolution based multiple signal classification (MUSIC) method in SAR image processing. The targets to be separated are modeled, and this modeled data is used to demonstrate the performance of algorithms.

• Proceedings of the KSRS Conference
• /
• 2007.10a
• /
• pp.258-261
• /
• 2007

Although there have been lack of studies using X-band SAR data particularly for forestry application as compared to C-, and L-band SAR data, it has a potential to distinguish tree species because most signals are backscattered on the top of canopy. This study aimed to compare signal characteristics of multi-band SAR data including X-band for classifying tree species. The data used for the study are SIR-C/X-SAR data (X-, C-, L-band) obtained on Oct. 3, 1994 over the forest area near Seoul, S. Korea. Thirty ground sample plots were collected per each tree species. Initial comparison of backscattering coefficients among three SAR bands shows that X-band data showed better separation of tree species than C- and L-band SAR data irrespective of polarization. The weak penetrating in canopy layer might be possible source of information for X-band data to be useful for the classification of forest species and cover type mapping.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.9 no.6
• /
• pp.155-161
• /
• 2009

Any classification process using SAR images presupposes the reduction of multiplicative speckle noise, since the variations caused by speckle make it extremely difficult to distinguish between neighboring classes within the feature space. This paper focus an argument of effective filter for preserving the weak boundaries by using the proposed method. To reduce speckle noise without blurring the edges of reconstructed image use wavelet-based sigma filter. As a result, the edge information of reconstructed image reduce blurring. Simulation results show that proposed method gives a better subjective quality than conventional methods for the speckle noise.