• Journal of Environmental Impact Assessment
• /
• v.20 no.2
• /
• pp.199-205
• /
• 2011

The Saemangeum embankment construction have changed the flowing on the topography of the coastal marine environment. However, the variety of ecological factors are changing from outside of Saemangeum embankment area. The ecosystem of various marine organisms have led to changes by sea surface temperature. The aim of this study is to monitoring of sea surface temperature(SST) changes were measured by using thermal infrared satellite imagery, MODIS and Landsat. The MODIS data have the high temporal resolution and Landsat satellite data with high spatial resolution was used for time series monitoring. The extracted informations from sea surface temperature changes were compared with the dyke to allow them inside and outside of Saemangeum embankment. The spatial extent of the spread of sea water were analyzed by SST using MODIS and Landsat thermal channel data. The difference of sea surface temperature between inland and offshore waters of Saemangeum embankment have changed by seasonal flow and residence time of sea water in dyke.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.23 no.9A
• /
• pp.2325-2332
• /
• 1998

In this paper, the quality scalable coding of selected region is presented. If a region is semantically more important than the others, it is appropriate that the image compression shcem is capable of handling the regional semantic difference because the information loss of the region of interest is more severe. We propose the quality scalable coding with its model by interoducing the quality scale parameter. It is more extended and generalized image compression philosophy than te conventional coding. As an implementation of the proposed quality scalable coding, H.263 based scheme is presented. This scheme can control the temporal and spatial quality efficiently, and improve the reconstructed image quality of the region of interest.

• Animal cells and systems
• /
• v.16 no.2
• /
• pp.135-144
• /
• 2012

$Dendronephthya$ $castanea$ Utinomi, 1952 is a member of the family Nephtheidae, and dominates shallow waters adjacent to the southern part of Jejudo Island, Korea. This species is a gonochoric internal brooder with a sex ratio of 1:1, and releases planulae around the time of the full and new moon from July to September, when the seawater temperature peaks. The gametogenic cycle is annual, and oogenesis (12 months) is longer than spermatogenesis (4-5 months). No difference in reproductive features including sexuality, sex ratio, gametogenesis and gametogenic cycles was found between the sympatric species $D.$ $castanea$ and $D.$ $gigantea$, and there was no temporal reproductive isolation. Investigation of the morphological taxonomy and molecular biology of these species indicates that they have very similar or identical traits, suggesting an absence of speciation and a need for taxonomic reclassification.

• Proceedings of the Korean Association of Geographic Inforamtion Studies Conference
• /
• 2004.03a
• /
• pp.203-208
• /
• 2004

NDVI (Normalized Difference Vegetation Index)는 식생의 건강상태 및 농작물 생산량 추정등에 효과적인 식생지수로, 20년 이상 축적된 MOAA NDVI data의 경우, 식생의 시기적, 계절적 변화탐지가 가능해져 이를 바탕으로 한 가뭄지수들이 개발되어 가뭄 모니터링에 사용되어지고 있다 지난 2001년, 한반도는 기상관측 이래 90년만의 강수량 최저치를 기록하여 전국적인 대 가뭄의 피해를 입었으며, 특히 북한은 유엔이 선정한 가뭄에 가장 취약한 국가로 그로 인한 식량난이 더욱 악화되고 있어 가뭄에 대한 정보는 필수적이라 할 수 있다. 이에 본 연구에서는 1994～2002년의 식물 생장기(growing season : 3～10월)동안 NDVI 10일 최대값 합성영상 (10-day maximum composite data)을 사용하여 남북한으로 나누어진 한반도를 대상으로 각각의 식생현황을 파악 및 비교하고, 산림, 농지, 도시지역별로 NDVI와 가뭄의 주원인인 강수량과의 상관관계로 그 효용성을 분석하였다. 그 결과, NDVI는 1～2개월 전 강수량의 영향이 가장 컸으며, 특히 농지지역에서의 상관계수가 높게 나타났다.

• Proceedings of the KIEE Conference
• /
• 2007.10a
• /
• pp.457-458
• /
• 2007

우리는 동영상에서 낮은 연산량으로 강인하게 움직이는 물체를 탐지하는 방법을 제안한다. 동영상에서 움직이는 물체를 탐지하기 위한 많은 방법들이 제안되었는데, 각각의 방법은 접근 방법에 따라 탐지 성능과 처리속도에 trade-off가 존재한다. 최근 폭넓게 사용되고 있는 가우시안 혼합모델을 이용한 배경모델생성 법의 경우 탐지성능은 우수하나 연산량이 많고, 차영상(Temporal difference)을 이용한 방법은 연산량은 적으나 노이즈에 민감하게 반응한다. 또한 알고리즘 특성상 탐지된 물체에 Hole과 Ghost가 발생하는 문제가 있다. 우리는 이러한 단점들을 극복하기 위해 이 두 가지 알고리즘을 효율적으로 결합하여 움직이는 물체를 탐지하였다. 실험은 사람, 차, 오토바이와 같이 실외환경에서 흔히 움직이는 물체로 탐지되는 요소들이 다양한 환경에서 실시하였으며, 실험한 결과 배경영역에서 발생하는 노이즈는 효과적으로 제거하면서 움직이는 물체를 탐지하였다.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.9
• /
• pp.135-143
• /
• 1996

In this paper, we popose a neural network that detects edges of moving objects in an image using a neural network of hierarchical spatial filter with orientation selectivity. We modify the temporal difference network by adding a self loop to each neuraon cell to reduce the problems of phantom edge detected by the neural network proposed by kwon yool et al.. The modified neural network alleviates the phantom edges of moving objects, and also can detect edges of miving objects even for the noisy input. By computer simulation with real images, the proposed neural netowrk can extract edges of different orientation efficiently and also can reduce the phantom edges of moving objects.

• Proceedings of the IEEK Conference
• /
• 2007.07a
• /
• pp.269-270
• /
• 2007

In previous work[1], each codeblock is coded using a context which is adaptively selected from three kinds of scanning directions after 3D DWT. But some subbands still have correlations among the coefficients in horizontal, vertical or temporal direction. In this paper, we propose a new 3D DWT-based video compression technique in which the difference of coefficients is calculated in one of three directions for every codeblocks and coded by a context adaptively selected for each codeblock. Experimental results show that the proposed compression technique outperforms measurably and visually compared to the conventional DWT-based techniques.

• Speech Sciences
• /
• v.10 no.1
• /
• pp.27-39
• /
• 2003

This study attempts to verify the reliability of automatically generated segment labels as compared to those obtained by conventional labelling by hand. First of all, an autolabeller is constructed using the standard HMM speech recognition technique. For evaluation, we compare the automatically generated labels with manually annotated labels for the same speech data. The comparison is performed by calculating the temporal difference between an autolabel boundary and its corresponding hand label boundary. When the mismatched duration between two labels falls within 10 msec, we consider the autolabel as correct. The results suggest that overall 78% of autolabels are correctly obtained. It is found that the boundary of obstruents is better aligned than that of sonorants and vowels. In case of stop sound classes, strong stops in manner-of-articulation wise and velar stops in place-of-articulation wise show better performance in boundary alignment. The result suggests that more phone-specific consideration is necessary to improve autosegmentation performance.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.3
• /
• pp.633-640
• /
• 1999

Reinforcement learning performs learning through interacting with trial-and-error in dynamic environment. Therefore, in dynamic environment, reinforcement learning method like Q-learning and TD(Temporal Difference)-learning are faster in learning than the conventional stochastic learning method. However, because many of the proposed reinforcement learning algorithms are given the reinforcement value only when the learning agent has reached its goal state, most of the reinforcement algorithms converge to the optimal solution too slowly. In this paper, we present COMREL(COMpressed REinforcement Learning) algorithm for finding the shortest path fast in a maze environment, select the candidate states that can guide the shortest path in compressed maze environment, and learn only the candidate states to find the shortest path. After comparing COMREL algorithm with the already existing Q-learning and Priortized Sweeping algorithm, we could see that the learning time shortened very much.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.1033-1035
• /
• 2003

This study attempts to analyze the potential of global scale NDVI data archive to monitor regional scale droughts. Ten-days maximum value NDVI composite data of the northeast Asia region were acquired for the growing seasons from 1993 to 2003. Two NDVI-derived drought indices (SVI, VCI), reported from previous studies, were applied to the study area. Although the SVI and VCI are mainly developed for monitoring the drought condition at the agriculture crop and grasslands, it turned out that they were also effective to reveal the drought condition over the temperate mixed forest. The drought symptom lasts at least one or two months even after the normal raining begins.