• Proceedings of the Korean Information Science Society Conference
• /
• 2010.06b
• /
• pp.279-282
• /
• 2010

최근 단일 고해상도 위성영상과 건물과 지형에 대한 정보를 가지고 있는 수치지도(DEM:Digital Elevation Model)를 이용하여 3차원 영상으로 구축하는 연구가 많이 이루어지고 있다. 3차원영상을 구축하기 위해서는 건물과 지형의 위치, 크기, 형상에 관한 정보가 필요하나 위성영상만으로는 구하기가 어려워 수치지도를 활용한다. 본 논문에서는 수치지도를 이용하여 3차원 건물과 지형의 고도 정보를 추출하고, 위성영상과 Snake 모델을 이용하여 반자동으로 추가적인 건물 높이와 바닥면에 대한 정보를 추출하여, 3차원 영상을 생성 하는 방법을 제안한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.11b
• /
• pp.901-903
• /
• 2005

본 논문은 스테레오 카메라를 이용한 이동 카메라 환경에서 다수의 사람을 검출하여 검출된 사람을 추적하는 방법을 제안한다. 카메라가 이동하게 되면 카메라의 움직임과 검출 대상이 되는 사람의 움직임이 동시에 발생하기 때문에 카메라 움직임을 변환 모델을 사용하여 보정하고, 독립적인 움직임을 추출하여 사람을 검출 하였다. 추적은 검출된 사람 영역의 컬러 분포에 기반하여 평균 이동(Mean Shift) 알고리즘을 적용하였다. 평균 이동 알고리즘은 빠르고 안정적인 성능으로 실시간 추적에 적합하다. 그러나 객체의 컬러 정보만으로는 배경과 컬러 분포가 유사한 객체의 경우 추적에 실패할 수 있는 단점이 있다. 이점을 보완하기 위하여 본 논문에서는 변위 지도(Disparity map)를 결합하여 객체와 배경을 분리하는 깊이 마스크를 생성하였다. 변위 지도를 사용하여 다수의 사람이 등장 할 경우 발생하는 가려짐, 겹침 등 다양한 실내 환경에서 발생하는 문제도 해결 하였다. 본 논문에서 제안하는 알고리즘은 다양한 데이터에 대해서 실험한 결과 정확한 검출과 추적에 우수한 성능을 확인 할 수 있었다.

• Korean Journal of Remote Sensing
• /
• v.12 no.1
• /
• pp.17-25
• /
• 1996

This paper presents a GCP marking technique using scanned maps. Twelve maps with a scale of 1:250, 000 were scanned and stored as raster images. The distortion factors of scanned maps were modelled by polynomials. The coefficients of the polynomials were determined by a least squre fit. This technique does not require a bulky and expensive digitizing table so that it is suitable for a low-cost pre-processing system. The GCP marking using this technique showed a sufficient accuracy for KITSAT1, 2 narrow camera images.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.01a
• /
• pp.47-48
• /
• 2019

본 논문에서는 신경망의 일종인 자기조직화지도(Self Organizing Map)을 이용하여 컨트롤러의 순서를 정하는 모델을 제안하였다. 자기조직화지도는 자율 학습에 의한 클러스터링을 수행하는 알고리즘으로써 컨트롤러에 가중치를 부여하고 컨트롤러 간 거리를 계산하여 효율적인 컨트롤러 선택을 목표로 한다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2021.06a
• /
• pp.297-300
• /
• 2021

영상 잡음 제거는 잡음으로 저하된 영상으로부터 잡음 없는 영상을 복원하는 기술이다. 최근 영상 처리에 딥러닝을 사용한 학습 기반 방법 중 저수준 컴퓨터 비전 분야에 고수준 영상 정보를 활용하는 접근이 있었다. 본 논문에서는 고수준 영상 정보인 영상 분할 지도를 활용하여 영상 속 가산 백색 잡음 제거 연구를 진행하였다. 잔차 연결을 활용한 구조의 인공신경망 모델에 잡음 영상, 잡음 수준 지도, 영상 분할 지도를 입력으로 넣어 고수준 영상 정보를 활용할 수 있게 하였다. 본 논문에서 제안한 인공신경망을 Outdoor Scene Dataset과 CBSD68 Dataset에 대해 확인해본 결과, PSNR과 인지적인 측면에서 DnCNN과 FFDNet보다 성능이 향상되는 것을 확인하였다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.3
• /
• pp.223-230
• /
• 2001

Since self-organizing map (SOM) preserves the topology of ordering in input spaces and trains itself by unsupervised algorithm, it is Llsed in many areas. However, SOM has a shortcoming: structure cannot be easily detcrmined without many trials-and-errors. Structure-adaptive self-orgnizing map (SASOM) which can adapt its structure as well as its weights overcome the shortcoming of self-organizing map: SASOM makes use of structure adaptation capability to place the nodes of prototype vectors into the pattern space accurately so as to make the decision boundmies as close to the class boundaries as possible. In this scheme, the initialization of weights of newly adapted nodes is important. This paper proposes a method which optimizes SASOM with genetic algorithm (GA) to determines the weight vector of newly split node. The leanling algorithm is a hybrid of unsupervised learning method and supervised learning method using LVQ algorithm. This proposed method not only shows higher performance than SASOM in terms of recognition rate and variation, but also preserves the topological order of input patterns well. Experiments with 2D pattern space data and handwritten digit database show that the proposed method is promising.

• Journal of the Korean earth science society
• /
• v.35 no.6
• /
• pp.412-421
• /
• 2014

In order to develop an effective mineral exploration technique, this study was carried out about the potential mapping of Gagok mine. The deposit model of Gagok mine is widely known. Based on the deposit model, we constructed mining indicator indices using related igneous rocks, faults, and carbonate rocks. By analyzing the spatial correlation between ore and indicator index structures, we decided the weighting values of indices according to the distance from the index structure. The 3D potential mapping was performed using 3D geological model and geological indices. The analyzed potential map verified that the locations and patterns of high potential regions of the results were well matched with those of the known ore bodies. Using the potential mapping results, we could effectively predict the location of a high potential area that has similar geological settings with ore.

• Journal of the Korean Geographical Society
• /
• v.37 no.2
• /
• pp.191-202
• /
• 2002

Characteristics and patterns of geographic features and human activities can be interpreted in terms of spatiality and temporality. The necessity to record the historical changes and the ability to reason in the real world has lead to a new field of research so called Integrated Spatio-Temporal analysis. The objective of this study is to investigate temporal maps for Spatio-temporal analysis, which have the integration functionality for visualizing spatiality and temporality of the geographic appearances and human activities. Land information is composed of spatial, attribute and temporal data and requires spatio-temporal representations. It is possible to visualize spatio-temporal variations with spatio-temporal databases and temporal map produced by integrated data models. This study constructs spatio-temporal model for temporal maps of land price variation analysis. Taking advantage of the spatio-temporal model proposed here, it is possible to visualize spatio-temporal variations with spatio-temporal database and temporal map. On a practical level, this study would be extended and utilized to various geographic features.

• Journal of Environmental Impact Assessment
• /
• v.25 no.6
• /
• pp.525-531
• /
• 2016

This study integrates the Universal Soil Loss Equation(USLE) with GIS method to assess the soil erosion for national land cover map between 2007 and 2014. The land cover change map and C factors of USLE were applied to the estimation of spatial distribution of sediment yield. However, they generated distinct results because of differences in their applied methods and calculation processes of national land cover map. To generate the USLE model, C factors of MOE(Ministry of Environment) were compared with soil erosion of Inje stadium development area at the Naerin watershed in Gangwon province to 2014. The several thematic maps of research area such as land cover map, topographic and soil maps, together with tabular precipitation data used for soil erosion calculation. The land cover change were carried with level-2 and high level land cover map of MOE and estimated maximum double of soil erosion.

• Journal of Korea Spatial Information System Society
• /
• v.10 no.1
• /
• pp.51-63
• /
• 2008

Previous studies for routing In USN (Ubiquitous Sensor Networks) appear to be unreliable due to the dependence on non-spatial data and the lack of map overlay analysis. Multiple spatial parameters and radio propagation modeling techniques were integrated to derive RSSI (Received Signal Strength Indicator) value between route nodes and produce a highly reliable path map. It was possible to identify area-wide patterns of USN route subject to many different Influences (e.g. the specific effects of radio blocking factors such as the visible area, road area, cell duplicated area, and building density), which cannot be acquired by traditional non-spatial modeling. The quantitative evidence concerning the USN route for individual cell as well as entire study area would be utilized as major tools to visualize paths in real-time and to select alternative path when failure or audition of route node occurs.