• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권2호
• /
• pp.775-792
• /
• 2015

Person re-identification is an important and challenging task in computer vision with numerous real world applications. Despite significant progress has been made in the past few years, person re-identification remains an unsolved problem. This paper presents a novel appearance-based approach to person re-identification. The approach exploits region covariance matrix and color histograms to capture the statistical properties and chromatic information of each object. Robustness against low resolution, viewpoint changes and pose variations is achieved by a novel signature, that is, the combination of Log Covariance Matrix feature and HSV histogram (LCMH). In order to further improve re-identification performance, third-party image sets are utilized as a common reference to sufficiently represent any image set with the same type. Distinctive and reliable features for a given image set are extracted through decision boundary between the specific set and a third-party image set supervised by max-margin criteria. This method enables the usage of an existing dataset to represent new image data without time-consuming data collection and annotation. Comparisons with state-of-the-art methods carried out on benchmark datasets demonstrate promising performance of our method.

• International Journal of Control, Automation, and Systems
• /
• 제3권3호
• /
• pp.461-468
• /
• 2005

This paper proposes a wheel-assembly automation system, which assembles a wheel into a hub of a vehicle hung to a moving hanger in a car manufacturing line. A macro-micro manipulator control strategy is introduced to increase the system bandwidth and tracking accuracy to ensure insertion tolerance. A camera is equipped at the newly designed wheel gripper, which is attached at the center of the end-effector of the macro-micro manipulator and is used to measure position error of the hub of the vehicle in real time. The redundancy problem in the macro-micro manipulator is solved without complicated calculation by assigning proper functions to each part so that the macro part tracks the velocity error while the micro part regulates the fine position error. Experimental results indicate that tracking error satisfies the insertion tolerance of assembly $({\pm}1mm)$, and thus it is verified that the proposed system can be applied to the wheel assembly task on a moving hanger in the manufacturing line.

• 한국게임학회 논문지
• /
• 제16권2호
• /
• pp.27-40
• /
• 2016

스텔스 게임에서 플레이어의 행동을 예측하는 것은 게임 디자인에 있어서 핵심적인 역할을 한다. 하지만, 플레이어와 게임 환경 간의 상호작용이 실시간으로 일어난다는 점에서 이러한 예측 프로세스를 자동화하는 것은 어려운 문제이다. 본 논문은 동적 환경에서의 스텔스 움직임을 예측하기 위한 강화학습 방법을 소개하며, 이를 위해 Q-learning과 인공신경망이 통합된 형태의 모델이 액션 시뮬레이션을 위한 분류기로 활용된다. 실험 결과들은 이러한 시뮬레이션 에이전트가 동적으로 변하는 주변 상황에 민감하게 반응함을 보여주며, 따라서 게임 레벨 디자이너가 다양한 게임 요소들을 결정하는데 유용함을 보여준다.

• International Journal of Aeronautical and Space Sciences
• /
• 제17권1호
• /
• pp.89-100
• /
• 2016

Given the unpredictability of the space environment, satellite communications are manually performed by exchanging telecommands and telemetry. Ground support for orbiting satellites is given only during limited periods of ground antenna visibility, which can result in conflicts when multiple satellites are present. This problem can be regarded as a scheduling problem of allocating antenna support (task) to limited visibility (resource). To mitigate unforeseen errors and costs associated with manual scheduling and mission planning, we propose a novel method based on a genetic algorithm to solve the ground support problem of multiple satellites and antennae with visibility conflicts. Numerous scheduling parameters, including user priority, emergency, profit, contact interval, support time, remaining resource, are considered to provide maximum benefit to users and real applications. The modeling and formulae are developed in accordance with the characteristics of satellite communication. To validate the proposed algorithm, 20 satellites and 3 ground antennae in the Korean peninsula are assumed and modeled using the satellite tool kit (STK). The proposed algorithm is applied to two operation modes: (i) telemetry, tracking, and command and (ii) payload. The results of the present study show near-optimal scheduling in both operation modes and demonstrate the applicability of the proposed algorithm to actual mission control systems.

• 한국근적외분광분석학회:학술대회논문집
• /
• 한국근적외분광분석학회 2001년도 NIR-2001
• /
• pp.1284-1284
• /
• 2001

The transesterification of molten ethylene/vinylacetate (EVA) copolymers by octanol as a reagent and sodium methoxide as a catalyst in an extruder has been monitored by on-line near infrared (NIR) spectroscopy. A total of 60 NIR spectra were acquired for 37 minutes with the last spectrum recorded 31 minutes after the addition of octanol and catalyst was stopped. The experimental spectra show strong baseline fluctuations which are corrected for by multiplicative scatter correction (MSC). The chemometric methods of orthogonal projection approach (OPA) and multivariate curve resolution (MCR) were used to resolve the spectra and to derive concentration profiles of the species. The detailed analysis reveals the absence of completely pure variables that leads to small errors in the calculation of pure spectra. The initial estimation of a concentration that is necessary as an input parameter for MCR also presents a non-trivial task. We obtained results that were not ideal but applicable for practical concentration control. They enable a fast monitoring of the process in real-time and resolve the spectra of the EVA copolymer and the ethylene/vinyl alcohol (EVAL) copolymer to be very close to the reference spectra. The chemometric methods used and the decomposed spectra are discussed in detail.

• Journal of Computing Science and Engineering
• /
• 제8권1호
• /
• pp.43-56
• /
• 2014

Semi-partitioned scheduling is a new approach for allocating tasks on multiprocessor platforms. By splitting some tasks between processors, semi-partitioned scheduling is used to improve processor utilization. In this paper, a new semi-partitioned scheduling algorithm called SS-DRM is proposed for multiprocessor platforms. The scheduling policy used in SS-DRM is based on the delayed rate monotonic algorithm, which is a modified version of the rate monotonic algorithm that can achieve higher processor utilization. This algorithm can safely schedule any system composed of two tasks with total utilization less than or equal to that on a single processor. First, it is formally proven that any task which is feasible under the rate monotonic algorithm will be feasible under the delayed rate monotonic algorithm as well. Then, the existing allocation method is extended to the delayed rate monotonic algorithm. After that, two improvements are proposed to achieve more processor utilization with the SS-DRM algorithm than with the rate monotonic algorithm. According to the simulation results, SS-DRM improves the scheduling performance compared with previous work in terms of processor utilization, the number of required processors, and the number of created subtasks.

• 한국방사성폐기물학회:학술대회논문집
• /
• 한국방사성폐기물학회 2004년도 Proceedings of the 4th Korea-China Joint Workshop on Nuclear Waste Management
• /
• pp.205-219
• /
• 2004

The remote operation of the Advanced Spent Fuel Conditioning Process (ACP) is analyzed by using the 3D graphic simulation tools. The ACP equipment operates in intense radiation fields as well as in a high temperature. Thus, the equipment should be designed in consideration of the remote handling and maintenance. As well as suitable remote handling and maintenance technology needs to be developed along with the design of the process concepts. To develop such remote operation technology, we developed the graphic simulator which provides the capability of verifying the remote operability of the ACP without fabrication of the process equipment. In other words, by applying virtual reality to the remote maintenance operation, a remote operation task can be simulated in the graphic simulator, not in the real environment. The graphic simulator will substantially reduce the cost of the development of the remote handling and maintenance procedure as well as the process equipment, while at the same time developing a remote maintenance concept that is more reliable, easier to implement, and easier to understand.

• 한국항공운항학회지
• /
• 제18권4호
• /
• pp.92-99
• /
• 2010

This study aimed to provide empirical evidence about expert performance approach in aviation field and the results suggested that the amount of experience(e.g. total flight hour) is necessary but not sufficient index of a pilot's expertise or superior performance. 43 pilots participated and completed a spatial span task and SA (situation awareness) tasks. To explore the factors predicting the performance in routine and non-routine situations, discriminant analysis was conducted. The results of discriminant analysis indicated that different variables are related with the performance in routine and non-routine situation. The factors predicting performance in routine situation were the spatial span scores and total flight hours. On the other hand, the factors predicting performance in non-routine situation were age and the qualification for instrument flying. In real world, total flight time which represents the quantity of experience has been frequently used to predict flight abilities and as an important index of expertise. The results of this study suggest that these kinds of factors have to be used cautiously to predict the performance in abnormal situation.

• 한국콘텐츠학회논문지
• /
• 제9권1호
• /
• pp.177-186
• /
• 2009

가상세계란 컴퓨터를 매개로 하여 2D 혹은 3D로 구현된 가상 환경 속에서 실시간으로 상호작용할 수 있는 영속적 공간으로 이 중 15세 이하의 사용자를 대상으로 하는 가상세계를 어린이용 가상세계라 한다. 가상세계의 교육적 장점으로는 구체적 경험, 현존감, 경험의 공유, 협동 가능성, 유연한 환경을 꼽을 수 있으며 여기에 게임이 가지고 있는 흥미 요소인 명확한 목표, 도전, 협동, 경쟁, 환상을 보탬으로써 교육효과를 높일 수 있다. 한편 가상세계에서의 교육이 효과적으로 이루어지기 위해서는 목표, 동기, 관계, 도움의 네 영역에 속하는 요소들을 두루 고려해야 한다. 이를 기준으로 하여 국내 어린이용 가상세계인 <노리스쿨>, <미스터 텅두>, <아즈로크>를 분석해 본 결과 명확한 목표 제시, 과제의 맥락화, 적절한 강화 제공은 잘 이루어지고 있었으나 준비성, 사회적 설득, 비계 설정 부분의 보완이 필요했다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.250-255
• /
• 2005

For navigation of a service robot, mapping and localization are very important. To estimate the robot pose, the map of the environment is required and it can be built by exploration or SLAM. Exploration is the fundamental task of guiding a robot autonomously during mapping such that it covers the entire environment with its sensors. In this paper, an efficient exploration scheme based on the position probability of the end nodes of a topological map is proposed. In this scheme, a topological map is constructed in real time using the thinning-based approach. The robot then updates the position probability of each end node maintaining its position at the current location based on the Bayesian update rule using the range data. From this probability, the robot can determine whether or not it needs to visit the specific end node to examine the environment around this node. Various experiments show that the proposed exploration scheme can perform exploration more efficiently than other schemes in that, in most cases, exploration for the entire environment can be completed without directly visiting everywhere in the environment.