• 융합신호처리학회논문지
• /
• 제6권1호
• /
• pp.15-22
• /
• 2005

본 논문은 CCD 칼라 영상을 이용하여 얼굴을 인식할 수 있는 방법을 제안한다. YCbCr 컬러모델에서 피부색에 대한 색상 정보와 적응적인 피부범위 확장을 통하여 얼굴후보영역을 추출하였다. 추출된 얼굴후보영역을 이용하여 곡선전개 방식의 초기곡선으로 사용하여 얼굴영역을 정확히 추출하였다. 얼굴의 특징점을 추출하기 위하여 얼굴영역에서 칼라정보를 이용한 Eye Map과 Mouth Map을 이용하였다. Log-polar변환의 중심점을 얻기 위하여 검출된 얼굴의 특징점을 이용하였다. 특징벡터를 추출하기 위하여 DCT, 웨이브렛 변환을 통하여 추출한 계수들을 이용하였다. 제안된 방법의 타당성을 검토하기 위하여 BP 학습알고리즘을 사용하는 신경망에서 얼굴인식을 수행하였다. 실험결과, 제안한 방법이 입력영상의 회전, 크기변화에 대하여 기존의 방법에 비하여 강인한 인식결과를 얻을 수 있었다.

• 정보처리학회논문지A
• /
• 제10A권3호
• /
• pp.215-224
• /
• 2003

인터넷 기반의 클러스터 시스템 환경에서 알고리즘의 이식성을 높이기 위해서는 네트워크의 특성 및 노드의 이질성에 따른 부하 불균형, 그리고 네트워크나 노드의 결함과 같은 다양한 수행환경의 변화에도 효과적으로 적응할 수 있어야 한다. 본 논문에서 제안하는 Expanded-WF 알고리즘은 Weighted Factoring 알고리즘을 기반으로 부하공유론 위하여 적응할당정책과 개선된 고정 분할 단위 알고리즘을 적용하고 결함허용을 위하여 작업을 중복 수행하는 기법을 적용한다. 적응할당정책으로는 느린 종노드의 작업을 빠른 종노드가 대신 수행하는 기법을 적용하였고, 개선된 고정 분할 단위 알고리즘은 네트워크의 통신시간과 계산시간을 겹치게 하는 것이다. 두 개의 네트워크 환경으로 구성된 이기종의 클러스터 환경에서 PVM을 이용한 행렬의 곱셈 프로그램으로 실험한 결과, 본 논문에서 제안한 알고리즘이 NOW 환경에서 효율적인 Send, GSS, Weighted Factoring 알고리즘보다 각각 55%, 63%, 그리고 20% 효율적임을 보였으며, 또한 결함허용도 가능함을 보였다.

• 대한교통학회지
• /
• 제15권1호
• /
• pp.175-192
• /
• 1997

첨단 교통정보 시스템(ATIS)의 핵심요소라고 할 수 있는 동적경로안내 시스템 (Dynamic Route Guidance System)은 운전자가 목적지에 도착하기까지 실시간 교통정보를 토대로 최적경로를 안내해줌으로써 날로 심화되고 있는 교통혼잡을 최소화 할 수 있으리라 기대를 모으고 있다. 특히 교통사고나 긴급 도로공사 등으로 인해 발생하는 돌발적 교통혼잡하에서는 DRGS의 역할이 더욱 커질 것으로 예상되고 있다. 본 논문은 돌발적 교통혼잡하에서 보다 효과적인 DRGS의 경로안내 전략을 수립하고 평가하는데 그 목적이 있다. 이를 위해 우선 하부구조기반 DRGS와 개인차량기반 DRGS의 장단점을 비교하고 시스템 아키텍쳐와 경로안내전략의 관계를 규명하였다. 또한 효율적인 경로안내를 위해 사용자평형 (User Equilibrium) 경로안내전략과 시스템 최적화(System Optimal) 경로안내 전략을 이상형교통망 (Idealized Network)을 통해 비교 분석하였다. 그리고 돌발적 교통 혼잡하에서 사용자평형 경로 안내를 사용할 경우 야기될 수 있는 Braess Paradox 문제와 시스템 최적경로안내를 사용할 경우 일어날 수 있는 사용자 호응도(User Compliance) 문제를 동시에 감안한 적응 형 경로안내 전략을 개발하였다. 이 방법은 위의 경로 안내 전략들이 가지고 있는 장단점을 상황에 따라 평가하여 경로안내 전략을 선택하는 과정을 수행시간을 절약하지 못할 것으로 평가되면 사용자 호응도를 고려하여 사용자 평형 전략을 선택하도록 하였다. 돌발적 교통 혼잡하에서 통행 시간을 동적으로 예측하기 위해서는 이산 확정적 대기행렬모형 (Discrete Deterministic Queueing Model)이 적용되었다. 한편, 적응형 전략의 효율성을 평가하기 위 해 이상형교통망과 실제 미국 Virginia 주의 Fairfax Country에 소재한 주간 고속도로 66번 과 인접 교통망을 대상으로 각종 돌발교통혼잡상황을 전제로 한 Traffic Simulation과 정보 제공 시나리오를 INTEGRATION Model을 사용하여 실행하였다. 그 결과 적응형전략이 단지 사용자평형 경로안내전략만 사용하는 경우에 비해 교통 혼잡도와 유고상황의 체류정도에 따라 3%에서 10%정도까지 전체통행시간을 절약할 수 있다는 결론을 얻었다.

• 정보과학회 논문지
• /
• 제42권9호
• /
• pp.1162-1174
• /
• 2015

오늘날 네트워크는 높은 대역폭과 높은 지연을 갖는 HBDP (High Bandwidth Delay Product) 네트워크의 특징을 보인다. 기존 TCP는 혼잡윈도우 크기의 느린 증가와 급격한 감소로 인하여 HBDP 네트워크에 부적절하다. 기존 TCP의 문제점을 해결하기 위해 연구된 TCP들은 손실기반 TCP와 지연기반 TCP로 구분한다. 대다수의 TCP는 기존 Slow Start 동작을 사용하며 오버슈트로 인한 대량의 패킷 손실을 초래한다. Congestion Avoidance 동작의 경우 손실기반 TCP는 대역폭 낭비와 RTT (Round Trip Time) 공정성 문제가 있으며 지연기반 TCP는 낮고 느린 대역폭 점유 문제가 있다. 제안하는 기법은 병목구간의 버퍼상태를 통해 혼잡제어를 함으로써 Slow Start와 Congestion Avoidance의 문제를 개선한다. 성능평가를 통해 HBDP 네트워크에서 제안하는 기법이 기존 TCP보다 향상된 성능을 보임을 확인하였다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2016년도 학술발표회
• /
• pp.204-204
• /
• 2016

Thailand had set the National Water Management Strategy which covered main six areas in the next 12 years, i.e., by priority: (1) water for household, (2) water for agricultural and industrial production, (3) water for flood and drought management, (4) water for quality issue, (5) water from forest conservation and soil erosion protection, (6) water resources management. However due to the climate change impact, there is a question for all strategies is whether to complete this mission under future climate change. If the impact affects our target, we have to clarify how to mitigate or to adapt with it. Vulnerability assessment was conducted under the framework of ADB's (with the parameters of exposure, sensitivity and adaptive capacity) and the assessments were classified into groups due to their different characteristic and the framework of the National Water Management Strategy, i.e., water supply (rural and urban), water for development (agriculture and others), water disasters (floods (flash, overflow), drought, water quality). The assessments identified the parameters concerned and weight factors used for each groups via expert group discussions and by using GIS mapping technology, the vulnerability maps were produced. The maps were verified with present water situation data (floods, drought, water quality). From the analysis result of this water resources management strategy, we found that 30% of all projects face the big impacts, 40% with low impact, and 30% for no impact. It is clear that water-related agencies have to carefully take care approximately 70% of future projects to meet water resources management strategy. It is recommended that additional issues should be addressed to mitigate the impact from climate risk on water resource management of the country, i.e., water resources management under new risk based on development scenarios, relationship with area-based problems, priority definition by viewpoints of risk, vulnerability (impact and occurrence probability in past and future), water management system in emergency case and water reserve system, use of information, knowledge and technology in management, network cooperation and exchange of experiences, knowledge, technique for sustainable development with mitigation and adaptation, education and communication systems in risk, new impact, and emergency-reserve system. These issues will be described and discussed.

• Journal of Information Processing Systems
• /
• 제15권3호
• /
• pp.593-603
• /
• 2019

In wearable healthcare systems, sensor devices can be deployed in places around the human body such as the stomach, back, arms, and legs. The sensors use tiny batteries, which have limited resources, and old sensor batteries must be replaced with new batteries. It is difficult to deploy sensor devices directly into the human body. Therefore, instead of replacing sensor batteries, increasing the lifetime of sensor devices is more efficient. A transmission power control (TPC) algorithm is a representative technique to increase the lifetime of sensor devices. Sensor devices using a TPC algorithm control their transmission power level (TPL) to reduce battery energy consumption. The TPC algorithm operates on a closed-loop mechanism that consists of two parts, such as sensor and sink devices. Most previous research considered only the sink part of devices in the closed-loop. If we consider both the sensor and sink parts of a closed-loop mechanism, sensor devices reduce energy consumption more than previous systems that only consider the sensor part. In this paper, we propose a new approach to consider both the sensor and sink as part of a closed-loop mechanism for efficient energy management of sensor devices. Our proposed approach judges the current channel condition based on the values of various body sensors. If the current channel is not optimal, sensor devices maintain their current TPL without communication to save the sensor's batteries. Otherwise, they find an optimal TPL. To compare performance with other TPC algorithms, we implemented a TPC algorithm and embedded it into sensor devices. Our experimental results show that our new algorithm is better than other TPC algorithms, such as linear, binary, hybrid, and ATPC.

• Nuclear Engineering and Technology
• /
• 제51권3호
• /
• pp.807-817
• /
• 2019

Airborne Gamma Ray Spectrometry (AGRS) with its important applications such as gathering radiation information of ground surface, geochemistry measuring of the abundance of Potassium, Thorium and Uranium in outer earth layer, environmental and nuclear site surveillance has a key role in the field of nuclear science and human life. The Broyden-Fletcher-Goldfarb-Shanno (BFGS), with its advanced numerical unconstrained nonlinear optimization in collaboration with Artificial Neural Networks (ANNs) provides a noteworthy opportunity for modern AGRS. In this study a new AGRS system empowered by ANN-BFGS has been proposed and evaluated on available empirical AGRS data. To that effect different architectures of adaptive ANN-BFGS were implemented for a sort of published experimental AGRS outputs. The selected approach among of various training methods, with its low iteration cost and nondiagonal scaling allocation is a new powerful algorithm for AGRS data due to its inherent stochastic properties. Experiments were performed by different architectures and trainings, the selected scheme achieved the smallest number of epochs, the minimum Mean Square Error (MSE) and the maximum performance in compare with different types of optimization strategies and algorithms. The proposed method is capable to be implemented on a cost effective and minimum electronic equipment to present its real-time process, which will let it to be used on board a light Unmanned Aerial Vehicle (UAV). The advanced adaptation properties and models of neural network, the training of stochastic process and its implementation on DSP outstands an affordable, reliable and low cost AGRS design. The main outcome of the study shows this method increases the quality of curvature information of AGRS data while cost of the algorithm is reduced in each iteration so the proposed ANN-BFGS is a trustworthy appropriate model for Gamma-ray data reconstruction and analysis based on advanced novel artificial intelligence systems.

• Journal of Astronomy and Space Sciences
• /
• 제35권4호
• /
• pp.253-261
• /
• 2018

Korea Astronomy and Space Science Institute (KASI) has been developing the space optical and laser tracking (SOLT) system for space geodesy, space situational awareness, and Korean space missions. The SOLT system comprises satellite laser ranging (SLR), adaptive optics (AO), and debris laser tracking (DLT) systems, which share numerous subsystems, such as an optical telescope and tracking mount. It is designed to be capable of laser ranging up to geosynchronous Earth orbit satellites with a laser retro-reflector array, space objects imaging brighter than magnitude 10, and laser tracking low Earth orbit space debris of uncooperative targets. For the realization of multiple functions in a novel configuration, the SOLT system employs a switching mirror that is installed inside the telescope pedestal and feeds the beam path to each system. The SLR and AO systems have already been established at the Geochang station, whereas the DLT system is currently under development and the AO system is being prepared for testing. In this study, the design and development of the SOLT system are addressed and the SLR data quality is evaluated compared to the International Laser Ranging Service (ILRS) tracking stations in terms of single-shot ranging precision. The analysis results indicate that the SLR system has a good ranging performance, to a few millimeters precision. Therefore, it is expected that the SLR system will not only play an important role as a member of the ILRS tracking network, but also contribute to future Korean space missions.

• 대한건축학회논문집:구조계
• /
• 제35권11호
• /
• pp.155-162
• /
• 2019

The construction industry is the highest safety accident causing industry as 28.55% portion of all industries' accidents in Korea. In particular, falling is the highest accidents type composed of 60.16% among the construction field accidents. Therefore, we analyzed the factors of major disaster affecting the fall accident and then derived feature importances by considering various variables. We used data collected from Korea Occupational Safety & Health Agency (KOSHA) for learning and predicting in the proposed model. We have an effort to predict the degree of occupational fall accidents by using the machine learning model, i.e., Adaboost, short for Adaptive Boosting. Adaboost is a machine learning meta-algorithm which can be used in conjunction with many other types of learning algorithms to improve performance. Decision trees were combined with AdaBoost in this model to predict and classify the degree of occupational fall accidents. HyOperpt was also used to optimize hyperparameters and to combine k-fold cross validation by hierarchy. We extracted and analyzed feature importances and affecting fall disaster by permutation technique. In this study, we verified the degree of fall accidents with predictive accuracy. The machine learning model was also confirmed to be applicable to the safety accident analysis in construction site. In the future, if the safety accident data is accumulated automatically in the network system using IoT(Internet of things) technology in real time in the construction site, it will be possible to analyze the factors and types of accidents according to the site conditions from the real time data.

• Nuclear Engineering and Technology
• /
• 제54권2호
• /
• pp.608-616
• /
• 2022

Power peaking factors (PPF) is an important parameter for safe and efficient reactor operation. There are several methods to calculate the PPF at TRIGA research reactors such as MCNP and TRIGLAV codes. However, these methods are time-consuming and required high specifications of a computer system. To overcome these limitations, artificial intelligence was introduced for parameter prediction. Previous studies applied the neural network method to predict the PPF, but the publications using the ANFIS method are not well developed yet. In this paper, the prediction of PPF using the ANFIS was conducted. Two input variables, control rod position, and neutron flux were collected while the PPF was calculated using TRIGLAV code as the data output. These input-output datasets were used for ANFIS model generation, training, and testing. In this study, four ANFIS model with two types of input space partitioning methods shows good predictive performances with R² values in the range of 96%-97%, reveals the strong relationship between the predicted and actual PPF values. The RMSE calculated also near zero. From this statistical analysis, it is proven that the ANFIS could predict the PPF accurately and can be used as an alternative method to develop a real-time monitoring system at TRIGA research reactors.