Go to the main menu
Skip to content
Go to bottom
REFERENCE LINKING PLATFORM OF KOREA S&T JOURNALS
> Journal Vol & Issue
Korean Journal of Remote Sensing
Journal Basic Information
Journal DOI :
The Korean Society of Remote Sensing
Editor in Chief :
Volume & Issues
Volume 12, Issue 3 - Dec 1996
Volume 12, Issue 2 - Sep 1996
Volume 12, Issue 1 - Jun 1996
Selecting the target year
An Object-based Stereo Matching Algorithm for Thermal IR Cloud Images
Shin, Dong-Seok ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 199~199
In this paper, the disparities of clouds are automatically determined from a pair of nadir and forward ATSR images through automatic stereo matching algorithms. An edge-based matching algorithm and an area-based matching algorithm, which have been widely used for the stereo matching of digital imagery, were applied to cloudy images. In the previous research, the edge-based algorithm showed a fast but inaccurate matching performance for cloud images. It is well known that the area-based algorithm gives reliable and accurate matching results. However, it is computationally intensive and it shows poor matching performance at disparity discontinuities. An object-based matching algorithm is suggested as a new approach to the stereo matching of IR cloud images. It determines cloud objects in an image using a segmentation technique. The temperature information in each object is used to select the best matching candidate region. It was found that the object-based matching algorithm was much faster and more accurate than the area-based algorithm in case of matching IR cloud images. It also overcomes the disadvantage of the area-based algorithm at disparity discontinuities. The automatic disparity determination can lead to the automatic determination of cloud height by using the navigation model of the ATSR.
Estimating Atmospheric Effect in Thermal Infrared Remote Sensing Using Single Channel Multiple Viewing Angle Data
Kim, Kwang-Eun ; Lee, Tai-Sup ; Oh, Sung-Nam ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 219~219
An explicit formulation was derived for quantitative estimation of atmospheric effect in thermal infrared remote sensing without any measurements of atmospheric parameters or ground truths. The algorithm calculates atmospheric effect using the single channel double viewing angle thermal infrared data. The atmospheric transmittance and radiance can be explicitly determined from the eigenvector of the covariance matrix of double viewing angle data. The experimental results, using the data acquired with AA3600 airborne multispectral scanner, showed that the maximum error of the estimated sea surface temperature could be reduced to less than 0.6
from about 3
before correction. Since none of the operational orbiting satellites offers the possibility of observing the same scene at different view angle within a short period of time, the algorithm is limited as a atmospheric correction technique for airborne data.
Mapping of Global Potential Water Resources using GIS
Ahn, Chung-Hyun ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 229~229
Global environmental data sets and CIS(Geographic Information System) techniques with simplified water balance model were employed to estimate water resources in land surfaces. Although there are more complex parameterizations of the hydrological components of the land-atmosphere interactions than are incorporated in this model, the modeled results shows that it has implications for future water resources and indicates possible areas of improvement in model parameterizations of surface processes. The more reliable data sets and hydrological models which can apply to global scale are necessary in order to estimate water resources more precisely. With the improvement of global data sets, the method employed in this study can readily be applied to various parts of the world for estimating water resources which have many practical applications. From the water budget analysis using developed data sets, one can identify the areal extent of water resources for each zone/region as well as total estimates of supply, demand and the resulting deficit or surplus of water resources, and thus defines critical areas and problems that may arise in the future. Therefore, the highest priority for future research must be placed on the development of global GIS software and preparation of worldwide information of hydrological components with better resolution and accuracy.
An Evaluation of the Green Belt of Seoul Metropolitan Region by Utilizing NDVI Derived from TM Data
Park, Chong-Hwa ; Suh, Dong-Jo ; Suh, Chang-Wan ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 245~245
The goal of this study was to investigate the effectiveness of the green belt of SMR(Seoul Metropolitan Region) for the protection of urban green spaces during recent new town construction boom by utilizing NDVI derived from 'N data. This research had two objectives. First, the protective function of the green belt for the green spaces of the study area were investigated by comparing NDVI of the green belt and that of the rest of the region. Second, NDVI change during recent new town construction boom of 1987∼1993 period was analyzed for different geographic conditions and ground cover types so as to investigate the impact of development on the green belt of the region. A change detection based on simple differencing of NDVI derived from TM data of the identical date of May20 of 1987 and 1993 were employed, and findings of this study can be summarized as follows. It was found that the average NDVI of pixels on the green belt was higher than that of areas not zoned for green belt, and average NDVI of pixels on the green belt was significantly improved for most topographic conditions and ground cover types during this period. Thus it can be concluded that the adverse impact of air pollution, acid rain, and development activities of the region on vegetation was negligible, and the most important objectives of green belt, which is to control urban sprawl, has been successfully achieved in spite of the rapid expansion of the SMR.
Applications of Principal Component Analysis and Fuzzy Set Operation to Change Detection of Urban Environment using Landsat Data
Lee, Ki-Won ; Park, Sung-Mi ; Chi, Kwang-Hoon ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 257~257
As for change detection research among several practical applications using remotely sensed datasets, various methodologies have been developed with computerized image processing techniques. Among them, Principal Component Analysis (PCA) and fuzzy set operation have been separately utilized to process image reduction and information extraction, if membership functions towards targets are given, from multiple images or raster datasets, respectively. Whilst, there are currently a few theoretical or practical studies to generalize these useful techniques towards image processing and interpretation of remotely sensed data; in this study, a generalized scheme by using PCA and fuzzy set operation is applied to urban-environmental change detection problems with using multi-spectral and multi-temporal Landsat MSS/TM imageries with an actual case study of Hanam city, nearby Seoul. In addition, three cases are applied to this methodology for practical aspects. differenced image, AGSM (Adaptive Grey Scale Mapping) image after differencing, and normalized differenced image. Among them, integrated images with AGSM image are more reasonable one than other cases, to some extents. On the resultant images, eight time-duration classes being capable of interpreting changed history are presented. Conclusively, it is thought that the generalized scheme used in this study is regarded as one of effective methodologies for urban-environmental change detection from satellite imageries.
Trophic State Mapping of Daechung Reservoir by Neural Network Classification of Landsat TM Images
Kim, Kwang-Eun ; Kim, Tae-Keun ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 271~271
This paper presents an application of a neural network to the trophic state mapping of Deachung reservoir. Landsat TM data acquired on 20 June, 1995 was used as an image data, and fifteen water sampling measurements were acquired during the satellite overpass. From the water sample measurements, trophic state of the fifteen sampling points were determined. These point data of trophic state and DN values were used as input pattern for training the network. A simple three layer feed forward network was trained by generalized delta rule using the input patterns of trophic state. It was possible for the trained artificial neural network to classify the image data accurately even though there were very small number of training data. The finally produced trophic state map of Deachung reservoir showed synoptic geographic coverage of water quality which can not be expected from the traditional water sampling surveys.
Model-based Simulation of Remotely Sensed Imagery
Jung, Myung-Hee ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 283~283
A hierarchical stochastic model to characterize processes observed in remotely sensed multispectral data has been developed with the goal of providing a general methodology for mu~ temporal landscape simulation models. The new model is based on a comprehensive stochastic representation of the scene. At the higher level, a Markov Random Field is employed to model the region process as a large scale scene characteristic. A fuzzy approach, integrated with the multiresolution data structure, is utilized to model the mixed and spectrally undefined cases around adjacent regions. At the lower level of the hierarchy, the model represents the natural variability within each region, parameterizing important characteristics of the continuous radiance field. Two approaches based on different assumptions are utilized for this. The first model is composed of two independent processes, one having a class dependent distribution, the other being a contaminating noise process. The second model employs a random field and includes possible contextual information. The new simulation model is evaluated using both simulated data and Landsat Thematic Mapper (TM) imagery.
Adaptive Area Based Matching Algorithm using Multi-windows and Multi-thresholds
Lee, Jae-Bum ; Jeon, Byung-Min ; Um, Gi-Mun ; Lee, Kwae-Hi ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 303~303
This paper describes an accurate DEM (Digital Elevation Model) generation method using a stereo matching from the SPOT satellite stereo images . The DEM generation process consists of satellite modeling, stereo matching using an image pair and height extraction. The matching algorithm has a great effect on the accuracy of the DEM generation algorithm. The matching strategy consists of two main sequential stages. The first stage is to find the search area from a priori knowledge about the geometry of a SPOT stereo model and the second stage is to find the corresponding point on both images. In this paper, we used a new area-based matching algorithm to find matched points with the normalized cross correlation as a similarity measure. A new area-based matching algorithm is to change the search window and target window for multi-threshold of correlation. Comparing with conventional matching algorithms using only one window and a threshold, the proposed algorithm used multi-window and multi-threshold. The input stereo images are
, level 1A panchromatic digital SPOT images of Chung-Chong Province, Korea, and our experimental results show that the number of matched points is much increased comparing with the conventional matching algorithm, while the number of mismatched points is not increased.
Unsupervised Image Segmentation for Spatially Continuous Imagery Using Gibbs Random Field
Lee, Sang-Hoon ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 317~317
An algorithm which makes use of spatial contextual information in a hierarchical clustering procedure has been developed for unsupervised image segmentation. Most environmental studies presume that a "surface patch" in an earth scene is likely to be spatially continuous and cohesive. The spatial smoothness is incorporated into the similarity measure that is established from a Bayesian function combining a discrete random field for region-class processes and a continuous random field for intensity processes in digital imagery. A Gibbs random field is used to quantify the spatial smoothness probabilistically. An GRF-based iterative algorithm is utilized to improve computational efficiency for the hierarchical clustering segmentation by providing an initial class configuration of the image with a moderate number of classes. The developed algorithm was applied to MonteCarlo simulation data and AVHRR data of the Korean Peninsula.
An Image Analysis Technique for Manganese Nodules On Deep Seabed Surfaces
Park, Seung-Ho ; Kim, Dae-Hee ; Kim, Choon-Woo ; Park, Chan-Young ; Kang, Jung-Keuk ;
Korean Journal of Remote Sensing, volume 12, issue 3, 1996, Pages 333~333
As the 7th pioneering investor for the future exploitation in the Clarion-Clipperton Fracture Zone in the northeastern pacific, Korea has been conducting the research cruises for more than 100 days per year since 1994. Faring the research cruises, the seabed surface is photographed every 30 seconds by the 35mm still camera. Features such as the area coverage and size distribution of the manganese nodules on the photographs serve as the essential information to determine the potential mining areas. This paper presents a method to calculate the features of the nodules using the digital image processing techniques. The proposed method consists of the following subtasks; the recognition of the data written on the photographs, compensation for non-uniform illumination, extraction of nodules, and calculation of the features. The experimental results with the sample seabed photographs indicate that the proposed method could be utilized as an efficient tool to process the massive photographs of the seabed surface.