• Korean Journal of Remote Sensing
• /
• v.37 no.4
• /
• pp.719-731
• /
• 2021

This study proposes a two-stage hybrid classification model for crop classification using multi-temporal remote sensing images; the model combines feature embedding by using an autoencoder (AE) with a convolutional neural network (CNN) classifier to fully utilize features including informative temporal and spatial signatures. Long short-term memory (LSTM)-based AE (LAE) is fine-tuned using class label information to extract latent features that contain less noise and useful temporal signatures. The CNN classifier is then applied to effectively account for the spatial characteristics of the extracted latent features. A crop classification experiment with multi-temporal unmanned aerial vehicle images is conducted to illustrate the potential application of the proposed hybrid model. The classification performance of the proposed model is compared with various combinations of conventional deep learning models (CNN, LSTM, and convolutional LSTM) and different inputs (original multi-temporal images and features from stacked AE). From the crop classification experiment, the best classification accuracy was achieved by the proposed model that utilized the latent features by fine-tuned LAE as input for the CNN classifier. The latent features that contain useful temporal signatures and are less noisy could increase the class separability between crops with similar spectral signatures, thereby leading to superior classification accuracy. The experimental results demonstrate the importance of effective feature extraction and the potential of the proposed classification model for crop classification using multi-temporal remote sensing images.

• Korean Journal of Remote Sensing
• /
• v.19 no.5
• /
• pp.381-392
• /
• 2003

Multi-temporal approaches using sequential data acquired over multiple years are essential for satisfactory discrimination between many land-cover classes whose signatures exhibit seasonal trends. At any particular time, the response of several classes may be indistinguishable. A harmonic model that can represent seasonal variability is characterized by four components: mean level, frequency, phase and amplitude. The trigonometric components of the harmonic function inherently contain temporal information about changes in land-cover characteristics. Using the estimates which are obtained from sequential images through spectral analysis, seasonal periodicity can be incorporates into multi-temporal classification. The Normalized Difference Vegetation Index (NDVI) was computed for one week composites of the Advanced Very High Resolution Radiometer (AVHRR) imagery over the Korean peninsula for 1996 ～ 2000 using a dynamic technique. Land-cover types were then classified both with the estimated harmonic components using an unsupervised classification approach based on a hierarchical clustering algorithm. The results of the classification using the harmonic components show that the new approach is potentially very effective for identifying land-cover types by the analysis of its multi-temporal behavior.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.30 no.1
• /
• pp.1-10
• /
• 2002

The purpose of this article is to examine land use change in the fringe area of a metropolitan city through multi-temporal data analysis. Change detection has been regarded as one of the most important applications for utilization of remotely sensed imageries. Conventionally, two images were used for change detection, and Arithmetic calculators were generally used on the process. Meanwhile, multi-temporal change detection for a large number of images has been carried out. In this paper, a digital land-use map and three Landsat TM data were utilized for the multi-temporal change detection Each urban area map was extracted as a base map on the process of multi-temporal change detection. Each urban area map was converted to bit image by using boolean logic. Various urban change types could be obtained by stacking the urban area maps derived from the multi-temporal data using Geographic Information System(GIS). Urban change type map was created by using the process of piling up the bit images. Then the urban change type map was compared with each land cover map for the change detection. Dalseo-gu of Daegu city and Hwawon-eup of Dalsung-gun, the fringe area of Daegu Metropolitan city, were selected for the test area of this multi-temporal change detection method. The districts are adjacent to each other. Dalseo-gu has been developed for 30 yeais and so a large area of paddy land has been changed into a built-up area. Hwawon-eup, near by Dalseo-gu, has been influenced by the urbanization of Dalseo-gu. From 1972 to 1999, 3,507.9ha of agricultural area has been changed into other land uses, while 72.7ha of forest area has been altered. This agricultural area was designated as a 'Semi-agricultural area'by the National landuse Management Law. And it was easy for the preserved area to be changed into a built-up area once it would be included as urban area. Finally, the method of treatment and management of the preserved area needs to be changed to prevent the destruction of paddy land by urban sprawl on the urban fringe.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.242-242
• /
• 2015

Monitoring the land cover changes in Nakdong River Basins using the multi-temporal remote sensing datasets is necessary for preserving properties in the river basins and monitoring the environmental changes in the river basins after the 4 major river restoration project. This research aims to monitor the land cover changes using the multi-temporal Landsat imageries and the airborne topographic LiDAR data. Firstly, the river basin boundaries are determined by using the LiDAR data, and the multiple river basin imageries are generated from the multi-temporal Landsat imageries by using the river basin boundaries. Next the classification method is employed to identify the multiple land covers in the generated river basin imageries. Finally, monitoring the land cover changes is implemented by comparing the differences of the same clusters in the multi-temporal river basin imageries.

• Korean Journal of Remote Sensing
• /
• v.35 no.4
• /
• pp.573-587
• /
• 2019

A multi-temporal approach using remotely sensed time series data obtained over multiple years is a very useful method for monitoring land covers and land-cover changes. While spectral-based methods at any particular time limits the application utility due to instability of the quality of data obtained at that time, the approach based on the temporal profile can produce more accurate results since data is analyzed from a long-term perspective rather than on one point in time. In this study, a multi-temporal approach applying a multi-periodic harmonic model is proposed for classification of remotely sensed data. A harmonic model characterizes the seasonal variation of a time series by four parameters: average level, frequency, phase, and amplitude. The availability of high-quality data is very important for multi-temporal analysis.An satellite image usually have many unobserved data and bad-quality data due to the influence of observation environment and sensing system, which impede the analysis and might possibly produce inaccurate results. Harmonic analysis is also very useful for real-time data reconstruction. Multi-periodic harmonic model is applied to the reconstructed data to classify land covers and monitor land-cover change by tracking the temporal profiles. The proposed method is tested with the MODIS and GOCI NDVI time series over the Korean Peninsula for 5 years from 2012 to 2016. The results show that the multi-periodic harmonic model has a great potential for classification of land-cover types and monitoring of land-cover changes through characterizing annual temporal dynamics.

• Korean Journal of Remote Sensing
• /
• v.18 no.1
• /
• pp.13-23
• /
• 2002

A radiometric correction method is developed to apply multi-temporal KOMPSAT-1 EOC satellite images for the detection of land-cover changes b\ulcorner recognizing changes in reflection pattern. Radiometric correction was carried out to eliminate the atmospheric effects that could interfere with the image properly of the satellite data acquired at different multi-times. Four invariant features of water, sand, paved road, and roofs of building are selected and a linear regression relationship among the control set images is used as a correction scheme. It is found that the utilization of panchromatic multi-temporal imagery requires the radiometric scene standardization process to correct radiometric errors that include atmospheric effects and digital image processing errors. Land-cover with specific change pattern such as paddy field is extracted by seasonal change recognition process.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_1
• /
• pp.999-1011
• /
• 2021

In spatio-temporal fusion aiming at predicting images with both high spatial and temporal resolutionsfrom multi-sensor images, the radiometric inconsistency between input multi-sensor images may affect prediction performance. This study investigates the effect of radiometric correction, which compensate different spectral responses of multi-sensor satellite images, on the spatio-temporal fusion results. The effect of relative radiometric correction of input images was quantitatively analyzed through the case studies using Sentinel-2, PlanetScope, and RapidEye images obtained from two croplands. Prediction performance was improved when radiometrically corrected multi-sensor images were used asinput. In particular, the improvement in prediction performance wassubstantial when the correlation between input images was relatively low. Prediction performance could be improved by transforming multi-sensor images with different spectral responses into images with similar spectral responses and high correlation. These results indicate that radiometric correction is required to improve prediction performance in spatio-temporal fusion of multi-sensor satellite images with low correlation.

• Korean Journal of Remote Sensing
• /
• v.21 no.2
• /
• pp.145-162
• /
• 2005

To improve the accuracy of land-cover discrimination in SAB data classification, this paper presents a methodology that includes feature extraction and fusion steps with multi-temporal SAR data. Three features including average backscattering coefficient, temporal variability and coherence are extracted from multi-temporal SAR data by considering the temporal behaviors of backscattering characteristics of SAR sensors. Dempster-Shafer theory of evidence(D-S theory) and fuzzy logic are applied to effectively integrate those features. Especially, a feature-driven heuristic approach to mass function assignment in D-S theory is applied and various fuzzy combination operators are tested in fuzzy logic fusion. As experimental results on a multi-temporal Radarsat-1 data set, the features considered in this paper could provide complementary information and thus effectively discriminated water, paddy and urban areas. However, it was difficult to discriminate forest and dry fields. From an information fusion methodological point of view, the D-S theory and fuzzy combination operators except the fuzzy Max and Algebraic Sum operators showed similar land-cover accuracy statistics.

• Korean Journal of Remote Sensing
• /
• v.31 no.4
• /
• pp.321-330
• /
• 2015

Atmospheric correction is an essential part in time-series analysis on biophysical parameters of surface features. In this study, we tried to examine possible problems in atmospheric correction of multitemporal High Spatial Resolution (HSR) images obtained from two different sensor systems. Three KOMPSAT-2 and two IKONOS-2 multispectral images were used. Three atmospheric correction methods were applied to derive surface reflectance: (1) Radiative Transfer (RT) - based absolute atmospheric correction method, (2) the Dark Object Subtraction (DOS) method, and (3) the Cosine Of the Uun zeniTh angle (COST) method. Atmospheric correction results were evaluated by comparing spectral reflectance values extracted from invariant targets and vegetation cover types. In overall, multi-temporal reflectance from five images obtained from January to December did not show consistent pattern in invariant targets and did not follow a typical profile of vegetation growth in forests and rice field. The multi-temporal reflectance values were different by sensor type and atmospheric correction methods. The inconsistent atmospheric correction results from these multi-temporal HSR images may be explained by several factors including unstable radiometric calibration coefficients for each sensor and wide range of sun and sensor geometry with the off-nadir viewing HSR images.

• Korean Journal of Remote Sensing
• /
• v.20 no.3
• /
• pp.163-173
• /
• 2004

The objective of this study is to develop a method to generate micro-relief digital elevation model (DEM) data of the tidal mudflats using multi-temporal Landsat Thematic Mapper (TM) data. Field spectroscopy measurements showed that reflectance of the exposed mudflat, shallow turbid water, and normal coastal water varied by TM band wavelength. Two sets of DEM data of the inter-tidal mudflat area were generated by interpolating several waterlines extracted from multi-temporal TM data acquired at different sea levels. The waterline appearing in the near-infrared band was different from the one in the middle-infrared band. It was found that the waterline in TM band 4 image was the boundary between the shallow turbid water and normal coastal water and used as a second contour line having 50cm water depth in the study area. DEM data generated by using both TM bands 4 and 5 rendered more detailed topographic relief as compared to the one made by using TM band 5 alone.