• IEIE Transactions on Smart Processing and Computing
• /
• 제6권1호
• /
• pp.53-59
• /
• 2017

Optical character recognition (OCR) automatically recognizes text in an image. OCR is still a challenging problem in computer vision. A successful solution to OCR has important device applications, such as text-to-speech conversion and automatic document classification. In this work, we analyze character recognition performance using the current state-of-the-art deep-learning structures. One is the AlexNet structure, another is the LeNet structure, and the other one is the SPNet structure. For this, we have built our own dataset that contains digits and upper- and lower-case characters. We experiment in the presence of salt-and-pepper noise or Gaussian noise, and report the performance comparison in terms of recognition error. Experimental results indicate by five-fold cross-validation that the SPNet structure (our approach) outperforms AlexNet and LeNet in recognition error.

• 소성∙가공
• /
• 제22권2호
• /
• pp.93-100
• /
• 2013

A tryout is a series of process optimization for robust stamping before transfer to the press shop. During tryout, the drawbead control, blank shape determination, binder surface modification, etc., are carried out mainly by a trial-and-error approach. As the level of difficulty of the stamping process increases, the formability becomes more sensitive to the contour of deformed shape, i.e. the blank shape. A digital tryout technique, which simulates a real tryout process, is proposed in this study for challenging stamping processes. Since digital tryout is carried out on a desktop, not in a press shop, a precise control of the deformed contour can be achieved if an optimal blank design technique is utilized. In this work, the proposed digital tryout technique is validated by successful applications to different automotive parts.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 1997년도 추계학술대회 논문집
• /
• pp.442-445
• /
• 1997

A challenging activity in the manufacturing industry is to perform in real time the continuous monitoring of the process state, the situation assessment and identification of the problem on line and diagnosis of the cause and importance of the problem if he process does not work properly. This paper describes LAT(Ladder Analysis Tool) system for fault tree generation to improving the fault diagnosis of CNC machine tools. The system consists of 4 steps which can automatically ladder analysis from ladder diagram to two diagnosis function models. The two diagnostic models based on he ladder diagram is switching function model and step switching function model. This system tries to overcome diagnosis deficiencies present machine tool.

• 한국터널지하공간학회 논문집
• /
• 제17권6호
• /
• pp.587-596
• /
• 2015

Recent years have shown great advances in the feasibility of long subsea tunnels. Projects such as the Channel Tunnel, the Stoerebelt Tunnel or the Bosporus Crossing have pushed the boundaries of TBM tunneling technology and fueled the work on feasibility studies for even more challenging projects such as crossing the Qiongzhou or Gibraltar Straits. There are numerous geotechnical challenges such as wide variations of ground conditions, high operation pressures or long tunnel distances and finally geological uncertainties which must be solved in order to attempt such projects. Several operational challenges such as large muck quantities interventions under difficult conditions and long transport distances also have to be tackled. TBM manufacturer and construction industry have developed a number of approaches to these challenges which point into the right technical direction and have been proven successfully in recent experiences. Their further development will allow attempting several megaprojects which are currently under discussion.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권6호
• /
• pp.2217-2229
• /
• 2015

Since the variations of illumination, the irregular changes of human shapes, and the partial occlusions, multiple person tracking is a challenging work in computer vision. In this paper, we propose a graph clustering method based on spatio-temporal information of moving objects for multiple person tracking. First, the part-based model is utilized to localize individual foreground regions in each frame. Then, we heuristically leverage the spatio-temporal constraints to generate a set of reliable tracklets. Finally, the graph shift method is applied to handle tracklet association problem and consequently generate the completed trajectory for individual object. The extensive comparison experiments demonstrate the superiority of the proposed method.

• 천문학회보
• /
• 제42권1호
• /
• pp.37.1-37.1
• /
• 2017

We consider a magnetic dipole model of a pulsar and investigate general relativistic effects on electromagnetic radiation from the pulsar. The general relativistic modifications should be found applicable to many well-known issues in pulsar astronomy. Among other things, the modifications of Goldreich-Julian model and subpulse drift would be of significant interest and challenging issues. The electromagnetic fields in the pulsar magnetosphere are computed by solving Maxwell's equations defined in the strongly curved spacetime around the pulsar, hence containing the properties of strong gravitational effects. On top of these effects, we also investigate the effects from rotation and obliqueness of the pulsar to work out the general relativistic versions of Goldreich-Julian model and subpulse drift.

• 한국세라믹학회지
• /
• 제40권7호
• /
• pp.625-631
• /
• 2003

Recent work at USC has focused on strategies to enhance the toughness and overall mechanical performance of polymer foams for use in lightweight sandwich structures. Both mechanical and chemical approaches have been employed with reasonable success. Fiber reinforcement, though difficult from a processing perspective, can lead to substantial enhancements in toughness and strength, while reducing friability. Chemical modifications are also challenging from a processing perspective, but can produce similar enhancements in performance. Efforts to enhance performance of phenolic foam and PVC foam through fiber reinforcement and chemical modification are described, along with the resulting enhancements in performance.

• 대한임베디드공학회논문지
• /
• 제11권4호
• /
• pp.219-225
• /
• 2016

Visual object tracking is one of the key tasks in computer vision. Robust trackers should address challenging issues such as fast motion, deformation, occlusion and so on. In this paper, we therefore propose a visual object tracking method that exploits inter-frame correlations of convolutional feature maps in Convolutional Neural Net (ConvNet). The proposed method predicts the location of a target by considering inter-frame spatial correlation between target location proposals in the present frame and its location in the previous frame. The experimental results show that the proposed algorithm outperforms the state-of-the-art work especially in hard-to-track sequences.

• Journal of Mechanical Science and Technology
• /
• 제18권5호
• /
• pp.789-797
• /
• 2004

The nanoscale sensing and manipulation have become a challenging issue in micro/nano-robotic applications. In particular, a feedback sensor-based manipulation is necessary for realizing an efficient and reliable handling of particles under uncertain environment in a micro/nano scale. This paper presents a piezoresistive MEMS cantilever for nanoscale force measurement in micro robotics. A piezoresistive MEMS cantilever enables sensing of gripping and contact forces in nanonewton resolution by measuring changes in the stress-induced electrical resistances. The calibration of a piezoresistive MEMS cantilever is experimentally carried out. In addition, as part of the work on nanomanipulation with a piezoresistive MEMS cantilever, the analysis on the interaction forces between a tip and a material, and the associated manipulation strategies are investigated. Experiments and simulations show that a piezoresistive MEMS cantilever integrated into a micro robotic system can be effectively used in nanoscale force measurements and a sensor-based manipulation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제4권4호
• /
• pp.452-474
• /
• 2010

Formidable growth of Internet technologies has revealed challenging issues about its scale and performance evaluation. Modeling and simulation play a central role in the evaluation of the behavior and performance of the large-scale network systems. Large numbers of nodes affect simulation performance, simulation execution time and scalability in a weighty manner. Most of the existing simulators have numerous problems such as size, lack of system theoretic approach and complexity of modeled network. In this work, a scalable discrete-event modeling approach is described for studying networks' scalability and performance traits. Key fundamental attributes of Internet and its protocols are incorporated into a set of simulation models developed using the Discrete Event System Specification (DEVS) approach. Large-scale network models are simulated and evaluated to show the benefits of the developed network models and approaches.