• ETRI Journal
• /
• 제42권3호
• /
• pp.411-419
• /
• 2020

In real-world intelligent transportation systems, accuracy in vehicle license plate detection and recognition is considered quite critical. Many algorithms have been proposed for still images, but their accuracy on actual videos is not satisfactory. This stems from several problematic conditions in videos, such as vehicle motion blur, variety in viewpoints, outliers, and the lack of publicly available video datasets. In this study, we focus on these challenges and propose a license plate detection and recognition scheme for videos based on a temporal matching prior network. Specifically, to improve the robustness of detection and recognition accuracy in the presence of motion blur and outliers, forward and bidirectional matching priors between consecutive frames are properly combined with layer structures specifically designed for plate detection. We also built our own video dataset for the deep training of the proposed network. During network training, we perform data augmentation based on image rotation to increase robustness regarding the various viewpoints in videos.

• 대한기계학회논문집
• /
• 제19권8호
• /
• pp.1907-1914
• /
• 1995

An on-line measurement of the workpiece hardened depth in laser surface hardening processes is very much difficult to achieve, since the hardening process occurs in depth wise direction. In this paper, the hardened depth is estimated using a multilayered neural network. Input data of the neural network are the surface temperatures at arbitrary chosen five surface points, laser power and traveling speed of laser beam torch. To simulate the actual hardening process, a finite difference method(FDM) is used to model the process. Since this model yields the calculation results of the temperature distribution around the workpiece volume in the vicinity of the laser torch, this model is used to obtain the network's training data and laser to evaluate the performance of the neural network estimator. The simulation results show that the proposed scheme can be used to estimate the hardened depth with reasonable accuracy.

• 한국컴퓨터산업교육학회:학술대회논문집
• /
• 한국컴퓨터산업교육학회 2003년도 제4회 종합학술대회 논문집
• /
• pp.119-126
• /
• 2003

Computer simulation has used to a area of military training from about several years ago. War game model(or computer simulation) endow a military man with field training such as combat experience without operating combat strength or capabilities. To samely construct simulation environment against actual combat environment is to well construct DB to operate war game model, associate among federates on network. Thus, we construct virtual combat environment enabling to efficiently manage network traffic among federates(or active nodes) on active network that construct virtual military training space such as urgent combat field needed to rapidly transfer combat information including image and video.

• Journal of Advanced Marine Engineering and Technology
• /
• 제39권8호
• /
• pp.863-869
• /
• 2015

This study analyzed the offshore plant industry and related regulations such as ISO, IEC, and Norsok Standards to develop an integrated network management system (INMS) capable of both on-site and remote management and configuration of IP-based network equipment in offshore plants. The INMS was designed based on actual specifications and POS plans, and a plan of management was verified through an offshore plant engineering company. Various modules such as PAGA interface modules, CCTV, IP-PBX, and HF-radio communication modules were developed for system implementation. Protocol and data design and screen design were followed by framework development and introduction of the automatic satellite communication function.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 하계학술대회 논문집 B
• /
• pp.401-403
• /
• 1998

This paper describes the neural network control method for the identification and control of chaotic nonlinear dynamical systems effectively. In our control method, the controlled system is modeled by an unknown NARMA model, and a feedforward neural network is used for identifying the chaotic system. The control signals are directly obtained by minimizing the difference between a setpoint and the output of the neural network model. Since learning algorithm guarantees that the output of the neural network model approaches that of the actual system, it is shown that the control signals obtained can also make the real system output close to the setpoint.

• 한국시뮬레이션학회논문지
• /
• 제9권2호
• /
• pp.103-118
• /
• 2000

ITS(Intelligent Transportation System) is an advanced system which can effectively handle the current transportation problems. DSRC(Dedicated Short Range Communication) is considered as a promising technology since it has the capability of two-way communication and can serves to implement various ITS services. In this paper, we study DSRC based ITS telecommunication traffic analysis and suggest an architecture and network design of telecommunication network for DSRC services. We also perform a simulation study to validate the proposed network architecture and design for Metropolitan Seoul Area with various network alternatives. In this simulation, we use actual traffic data and road characteristics from Seoul area and use our DSRC service configuration.

• 대한안전경영과학회지
• /
• 제17권3호
• /
• pp.265-272
• /
• 2015

The competition between companies for prior occupation of the market is becoming fierce. In this highly competitive situation, it is important for companies to differentiate themselves if they are going to have a chance at success. And the competition to create the best solution method possible is higher than ever. Increased competition is forcing companies to lower costs and improve efficiency. A supply chain management(SCM) has become one of the most important solution methods of competitive advantage. This study has developed a simulator for the supply chain network problem. The simulator is designed to simulate the conditions of an actual supply chain network considering uncertainties. The simulator developed using commercial simulation tool ARENA and the results of computational experiments for a simple example were given and discussed to validate the developed simulator. Further research is needed, but using the simulator could become a useful tool for decision making in the supply chain network area.

• Journal of Information Processing Systems
• /
• 제17권2호
• /
• pp.385-398
• /
• 2021

The complexity of deep learning models affects the real-time performance of gesture recognition, thereby limiting the application of gesture recognition algorithms in actual scenarios. Hence, a residual learning neural network based on a deep convolutional neural network is proposed. First, small convolution kernels are used to extract the local details of gesture images. Subsequently, a shallow residual structure is built to share weights, thereby avoiding gradient disappearance or gradient explosion as the network layer deepens; consequently, the difficulty of model optimisation is simplified. Additional convolutional neural networks are used to accelerate the refinement of deep abstract features based on the spatial importance of the gesture feature distribution. Finally, a fully connected cascade softmax classifier is used to complete the gesture recognition. Compared with the dense connection multiplexing feature information network, the proposed algorithm is optimised in feature multiplexing to avoid performance fluctuations caused by feature redundancy. Experimental results from the ISOGD gesture dataset and Gesture dataset prove that the proposed algorithm affords a fast convergence speed and high accuracy.

• International Journal of Computer Science & Network Security
• /
• 제22권3호
• /
• pp.285-291
• /
• 2022

WSN is the major component for information transfer in IoT environments. Source Location Privacy (SLP) has attracted attention in WSN environments. Effective SLP can avoid adversaries to backtrack and capture source nodes. This work presents a Two-Level Randomized Sector-based Routing (TLRSR) model to ensure SLP in wireless environments. Sector creation is the initial process, where the nodes in the network are grouped into defined sectors. The first level routing process identifies sector-based route to the destination node, which is performed by Ant Colony Optimization (ACO). The second level performs route extraction, which identifies the actual nodes for transmission. The route extraction is randomized and is performed using Simulated Annealing. This process is distributed between the nodes, hence ensures even charge depletion across the network. Randomized node selection process ensures SLP and also avoids depletion of certain specific nodes, resulting in increased network lifetime. Experiments and comparisons indicate faster route detection and optimal paths by the TLRSR model.

• 로봇학회논문지
• /
• 제18권1호
• /
• pp.88-92
• /
• 2023

Lithium-ion batteries are actively used in various industrial sites such as field robots, drones, and electric vehicles due to their high energy efficiency, light weight, long life span, and low self-discharge rate. When using a lithium-ion battery in a field, it is important to accurately estimate the SoC (State of Charge) of batteries to prevent damage. In recent years, SoC estimation using data-based artificial neural networks has been in the spotlight, but it has been difficult to deploy in the embedded board environment at the actual site because the computation is heavy and complex. To solve this problem, neural network lightening technologies such as network pruning have recently attracted attention. When pruning a neural network, the performance varies depending on which layer and how much pruning is performed. In this paper, we introduce an optimized pruning technique by improving the existing pruning method, and perform a comparative experiment to analyze the results.