• Proceedings of the KSRS Conference
• /
• 2004.10a
• /
• pp.10-13
• /
• 2004

Hyperspectral data has a great advantage to classify various surface materials that are spectrally similar. In this study, we attempted to classify man-made materials in urban area using Hyperion data. Hyperion imagery of Seoul was initially processed to minimize radiometric distortions caused by sensor and atmosphere. Using color aerial photographs. we defined seven man-made surfaces (concrete, asphalt road. railroad, buildings, roof, soil, shadow) for the classification in Seoul. The hyperspectral data showed the potential to identify those manmade materials that were difficult to be classified by multispectral data. However. the classification of road and buildings was not quite satisfactory due to the relatively low spatial resolution of Hyperion image. Further, the low radiometric quality of Hyperion sensor was another limitation for the application in urban area.

• Korean Journal of Remote Sensing
• /
• v.36 no.6_1
• /
• pp.1379-1391
• /
• 2020

UAV in the agricultural application are capable of collecting ultra-high resolution image. It is possible to obtain timeliness images for phenological phases of the crop. However, the UAV uses a variety of sensors and multi-temporal images according to the environment. Therefore, it is essential to use normalized image data for time series image application for crop monitoring. This study analyzed the variability of UAV reflectance and vegetation index according to Aviation Image Making Environment to utilize the UAV multispectral image for agricultural monitoring time series. The variability of the reflectance according to environmental factors such as altitude, direction, time, and cloud was very large, ranging from 8% to 11%, but the vegetation index variability was stable, ranging from 1% to 5%. This phenomenon is believed to have various causes such as the characteristics of the UAV multispectral sensor and the normalization of the post-processing program. In order to utilize the time series of unmanned aerial vehicles, it is recommended to use the same ratio function as the vegetation index, and it is recommended to minimize the variability of time series images by setting the same time, altitude and direction as possible.

• Proceedings of the KSRS Conference
• /
• 2002.10a
• /
• pp.65-65
• /
• 2002

In remote sensing, images are acquired over the same area by sensors of different spectral ranges (from the visible to the microwave) and/or with different number, position, and width of spectral bands. These images are generally partially redundant, as they represent the same scene, and partially complementary. For many applications of image classification, the information provided by a single sensor is often incomplete or imprecise resulting in misclassification. Fusion with redundant data can draw more consistent inferences for the interpretation of the scene, and can then improve classification accuracy. The common approach to the classification of multisensor data as a data fusion scheme at pixel level is to concatenate the data into one vector as if they were measurements from a single sensor. The multiband data acquired by a single multispectral sensor or by two or more different sensors are not completely independent, and a certain degree of informative overlap may exist between the observation spaces of the different bands. This dependence may make the data less informative and should be properly modeled in the analysis so that its effect can be eliminated. For modeling and eliminating the effect of such dependence, this study employs a strategy using self and conditional information variation measures. The self information variation reflects the self certainty of the individual bands, while the conditional information variation reflects the degree of dependence of the different bands. One data set might be very less reliable than others in the analysis and even exacerbate the classification results. The unreliable data set should be excluded in the analysis. To account for this, the self information variation is utilized to measure the degrees of reliability. The team of positively dependent bands can gather more information jointly than the team of independent ones. But, when bands are negatively dependent, the combined analysis of these bands may give worse information. Using the conditional information variation measure, the multiband data are split into two or more subsets according the dependence between the bands. Each subsets are classified separately, and a data fusion scheme at decision level is applied to integrate the individual classification results. In this study. a two-level algorithm using hierarchical clustering procedure is used for unsupervised image classification. Hierarchical clustering algorithm is based on similarity measures between all pairs of candidates being considered for merging. In the first level, the image is partitioned as any number of regions which are sets of spatially contiguous pixels so that no union of adjacent regions is statistically uniform. The regions resulted from the low level are clustered into a parsimonious number of groups according to their statistical characteristics. The algorithm has been applied to satellite multispectral data and airbone SAR data.

• Korean Journal of Remote Sensing
• /
• v.28 no.3
• /
• pp.277-285
• /
• 2012

The GOCI imagery can be an effective alternative to monitor short-term changes over terrestrial environments. This study aimed to assess the radiometric characteristics of the GOCI multispectral imagery for land applications. As an initial approach, we compared GOCI at-sensor radiance with MODIS data obtained simultaneously. Dynamic range of GOCI radiance was larger than MODIS over land area. Further, the at-sensor radiance over various land surface targets were tested by vicarious calibration. Surface reflectance were directly measured in field using a portable spectrometer and indirectly derived from the atmospherically corrected MODIS product over relatively homogeneous sites of desert, tidal flat, bare soil, and fallow crop fields. The GOCI radiance values were then simulated by radiative transfer model (6S). In overall, simulated radiance were very similar to the actual radiance extracted from GOCI data. Normalized difference vegetation index (NDVI) calculated from the GOCI bands 5 and 8 shows very close relationship with MODIS NDVI. In this study, the GOCI imagery has shown appropriate radiometric quality to be used for various land applications. Further works are needed to derive surface reflectance over land area after atmospheric correction.

• Korean Journal of Remote Sensing
• /
• v.29 no.4
• /
• pp.399-406
• /
• 2013

The quality of satellite imagery should be improved and stabilized to satisfy numerous users. The radiometric characteristics of an optical sensor can be a measure of data quality. In this study, a band aggregation technique and spectral response function of hyperspectral images are used to simulate multispectral images. EO-1 Hyperion and Landsat-8 OLI images acquired with about 30 minutes difference in overpass time were exploited to evaluate radiometric coefficients of OLI. Radiance values of the OLI and the simulated OLI were compared over three subsets covered by different land types. As a result, the index of agreement shows over 0.99 for all VNIR bands although there are errors caused by space/time and sensors.

• Korean Journal of Remote Sensing
• /
• v.33 no.5_1
• /
• pp.537-546
• /
• 2017

This study aims to develop the floating debris detection algorithm using a Unmanned Aerial Vehicle (UAV) and multispectral sensors. In addition, the occurrence range of floating debris was estimated by applying the algorithm. An aerial photograph using an unmanned aerial vehicle was used to generate an orthoimage that can calculate the area. A spectrum survey of water, plants litter, polystyrene foam etc. was conducted. After obtaining spectroscopic characteristics of floating debris and water, the River Floating Debris (RFD) index was calculated. And we detected the floating debris through band combination of sensor using RFD. As a result of the RFD application, accumulation zone of floating debris was confirmed at three sites in the orthoimage. It was estimated that a lot of floating debris was accumulated at 0.82 ha ($8,200m^2$), which is corresponding to 3.6% including the accumulation zone.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.29 no.2
• /
• pp.175-181
• /
• 2011

In remote sensing, it has been a difficult task to obtain a multispectral image with high spatial resolution because of the technical limitation of satellite sensors. In order to solve these problems, various pansharpening algorithms have been tried and proposed. However, most pansharpened images created by various approaches tend to distort the spectral characteristics of the original multispectral image or decrease the visual sharpness of the panchromatic image. To minimize the spectral distortion of pansharpened image while preserving spatial information of the panchromatic image, a hybrid pansharpening algorithm based on the spatial correlation was proposed. In this paper, we analyzed the spatial and spectral distortion of the hybrid pansharpened images generated by the various spatial correlation methods. In the experiments, we proved that the method by using Laplacian filtering was more efficient than other high frequency extraction algorithms in the viewpoint of spectral distortion and spatial sharpness.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.22 no.4
• /
• pp.86-101
• /
• 2019

Remote sensing technology has been continuously developed both quantitatively and qualitatively, including platform development, exploration area, and exploration functions. Recently, the use cases and related researches in the agricultural field are increasing. Also, since it is possible to detect and quantify the condition of cropland and establish management plans and policy support for cropland and agricultural environment, it is being studied in various fields such as crop growth abnormality determination and crop estimation based on time series information. The purpose of this study was to analyze the vegetation index for agricultural land reclamation area using a UAV equipped with a multi-spectral sensor. In addition, field surveys were conducted to evaluate the accuracy of vegetation indices calculated from multispectral image data obtained using UAV. The most appropriate vegetation index was derived by evaluating the correlation between vegetation index calculated by field survey and vegetation index calculated from UAV multispectral image, and was used to analyze vegetation index of the entire area.

• Proceedings of the KSRS Conference
• /
• 1998.09a
• /
• pp.409-414
• /
• 1998

To examine the detectability of the aerosol and/or Yellow dust from China crossing over the Yellow sea, three works carried out as follows , Firstly, a comparison was made of the visible(VIS), water vapor(WV), and Infrared(IR) images of the GMS-5 and NOAA/AVHRR on the cases of yellow sand event over Korea. Secondly, the spectral radiance and reflectance(%) was observed during the yellow sand phenomena on April, 1998 in Seoul using the GER-2600 spectroradiometer, which observed the reflected radiance from 350 to 2500 nm in the atmosphere. We selected the optimum wavelength for detecting of the yellow sand from this observation, considering the effects of atmospheric absorption. Finally, the atmospheric radiance emerging from the LOWTRAN-7 radiative transfer model was simulated with and without yellow sand, where we used the estimated aerosol column optical depth ($\tau$ 673 nm) in the Meteorological Research Institute and the d'Almeida's statistical atmospheric aerosol radiative characteristics. The image analysis showed that it was very difficult to detect the yellow sand region only by the image processing because the albedo characteristics of the sand vary irregularly according to the density, size, components and depth of the yellow sand clouds. We found that the 670-680 nm band was useful to simulate aerosol characteristics considering the absorption band from the radiance observation. We are now processing the simulation of atmospheric radiance distribution in the range of 400-900 nm. The purpose of this study is to present the preliminary results of the aerosol and/or Yellow dust detectability using the Ocean Scanning Multispectral Imager(OSMI), which will be mounted on KOMPSAT-1 as the ocean color monitoring sensor with the range of 400-900 nm wavelength.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2022.10a
• /
• pp.88-88
• /
• 2022

Detection of stress responses in crops is important to diagnose crop growth and evaluate yield. Also, the multi-spectral sensor is effectively known to evaluate stress caused by nutrient and moisture in crops or biological agents such as weeds or diseases. Therefore, in this experiment, multispectral images were taken by an unmanned aerial vehicle(UAV) under field condition. The experiment was conducted in the long-term fertilizer field in the National Institute of Crop Science, and experiment area was divided into different status of NPK(Control, N-deficiency, P-deficiency, K-deficiency, Non-fertilizer). Total 11 vegetation indices were created with RGB and NIR reflectance values using python. Variations in nutrient content in plants affect the amount of light reflected or absorbed for each wavelength band. Therefore, the objective of this experiment was to evaluate vegetation indices derived from multispectral reflectance data as input into machine learning algorithm for the classification of nutritional deficiency in rice. RandomForest model was used as a representative ensemble model, and parameters were adjusted through hyperparameter tuning such as RandomSearchCV. As a result, training accuracy was 0.95 and test accuracy was 0.80, and IPCA, NDRE, and EVI were included in the top three indices for feature importance. Also, precision, recall, and f1-score, which are indicators for evaluating the performance of the classification model, showed a distribution of 0.7-0.9 for each class.