• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.34 no.4
• /
• pp.349-356
• /
• 2016

This study investigates two camera self-calibration approaches, on-site self-calibration and laboratorial self-calibration, both of which are based on self-calibration theory and implemented by using a commercial photogrammetric solution, Agisoft PhotoScan. On-site self-calibration implements camera self-calibration and aerial triangulation by using the same aerial photos. Laboratorial self-calibration implements camera self-calibration by using photos captured onto a patterned target displayed on a digital panel, then conducts aerial triangulation by using the aerial photos. Aerial photos are captured by an unmanned aerial vehicle, and target photos are captured onto a 27in LCD monitor and a 47in LCD TV in two experiments. Calibration parameters are estimated by the two approaches and errors of aerial triangulation are analyzed. Results reveal that on-site self-calibration excels laboratorial self-calibration in terms of vertical accuracy. By contrast, laboratorial self-calibration obtains better horizontal accuracy if photos are captured at a greater distance from the target by using a larger display panel.

• Journal of Korea Multimedia Society
• /
• v.18 no.12
• /
• pp.1432-1438
• /
• 2015

We present a methodology to detect moving shadows in video sequences, which is considered as a challenging and critical problem in the most visual surveillance systems since 1980s. While most previous moving shadow detection methods used hand-crafted features such as chromaticity, physical properties, geometry, or combination thereof, our method can automatically learn features to classify whether image segments are shadow or foreground by using a deep learning architecture. Furthermore, applying Markov Random Field enables our system to refine our shadow detection results to improve its performance. Our algorithm is applied to five different challenging datasets of moving shadow detection, and its performance is comparable to that of state-of-the-art approaches.

• Journal of Digital Convergence
• /
• v.13 no.1
• /
• pp.227-235
• /
• 2015

Big data analysis is a means of organizational problem solving. For an effective problem solving, approaches to problem solving should take into account the factors such as characteristics of problem, types and availability of data, data analytic capability, and technical capability. In this article we propose three approaches: logical top-down, data driven bottom-up, and prototyping for overcoming undefined problem circumstances. In particular we look into the relationship of creative problem solving with the bottom-up approach. Based on the organizational data governance and data analytic capability, we also derive strategic issues concerning the sourcing of big data analysis.

• Journal of Institute of Control, Robotics and Systems
• /
• v.9 no.11
• /
• pp.917-924
• /
• 2003

Recurrent neural networks have been successfully applied to communications channel equalization. Major disadvantages of gradient-based learning algorithms commonly employed to train recurrent neural networks are slow convergence rates and long training sequences required for satisfactory performance. In a high-speed communications system, fast convergence speed and short training symbols are essential. We propose decision feedback equalizers using a recurrent neural network trained with Kalman filtering algorithms. The main features of the proposed recurrent neural equalizers, utilizing extended Kalman filter (EKF) and unscented Kalman filter (UKF), are fast convergence rates and good performance using relatively short training symbols. Experimental results for two time-varying channels are presented to evaluate the performance of the proposed approaches over a conventional recurrent neural equalizer.

• ETRI Journal
• /
• v.34 no.4
• /
• pp.621-624
• /
• 2012

A new miniaturized 4-port waveguide generating a transverse electromagnetic wave is proposed. The waveguide presents enhanced performance of higher field uniformity in extended test volume up to an increased test frequency limit compared to that of the conventional 2-port waveguide. The advantageous features of the proposed waveguide have been obtained through a new design scheme based on effective miniaturization maintaining good impedance matching. Consequently, we can provide a more accurate electromagnetic compatibility test method, covering larger devices operating in higher frequencies, which is a marked improvement upon the conventional approaches.

• Structural Engineering and Mechanics
• /
• v.31 no.4
• /
• pp.383-392
• /
• 2009

The static and dynamic analyses of simply supported beams are studied by using the U-transformation method and the finite difference method. When the beam is divided into the mesh of equal elements, the mesh may be treated as a periodic structure. After an equivalent cyclic periodic system is established, the difference governing equation for such an equivalent system can be uncoupled by applying the U-transformation. Therefore, a set of single-degree-of-freedom equations is formed. These equations can be used to obtain exact analytical solutions of the deflections, bending moments, buckling loads, natural frequencies and dynamic responses of the beam subjected to particular loads or excitations. When the number of elements approaches to infinity, the exact error expression and the exact convergence rates of the difference solutions are obtained. These exact results cannot be easily derived if other methods are used instead.

• Journal of Radiopharmaceuticals and Molecular Probes
• /
• v.8 no.1
• /
• pp.39-44
• /
• 2022

The development of methods for the inert and stable radiohalogenation of targeted radiopharmaceuticals is a prerequisite for real-time diagnosis and therapy using radiohalogenated radiopharmaceuticals. Radiohalogenated carboranes demonstrate superior stability in vivo and versatile applications compared with directly labeled tyrosine analogues or synthetically modified organic compounds. Herein, we focus on the most common approaches for the radioiodination (¹²³l, ¹²⁴l, ¹²⁵l, and ¹³¹l) of carborane derivatives.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.493-495
• /
• 2023

Deep learning methods and attention mechanisms have been incorporated to improve facial emotion recognition, which has recently attracted much attention. The fusion approaches have improved accuracy by combining various types of information. This research proposes a fusion network with self-attention and local attention mechanisms. It uses a multi-layer perceptron network. The network extracts distinguishing characteristics from facial images using pre-trained models on RAF-DB dataset. We outperform the other fusion methods on RAD-DB dataset with impressive results.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.805-807
• /
• 2021

In-hospital cardiac arrest is a significant problem for medical systems. Although the traditional early warning systems have been widely applied, they still contain many drawbacks, such as the high false warning rate and low sensitivity. This paper proposed a strategy that involves a deep learning approach based on a novel interpretable deep tabular data learning architecture, named TabNet, for the Rapid Response System. This study has been processed and validated on a dataset collected from two hospitals of Chonnam National University, Korea, in over 10 years. The learning metrics used for the experiment are the area under the receiver operating characteristic curve score (AUROC) and the area under the precision-recall curve score (AUPRC). The experiment on a large real-time dataset shows that our method improves compared to other machine learning-based approaches.

• Journal of Digital Convergence
• /
• v.14 no.3
• /
• pp.219-226
• /
• 2016

The aim of this study is to suggest the strategic implications for manufacturing 3.0 in Korea by reviewing an innovation approaches of German that is a source of manufacturing innovation in Europe. Today, growth potential of korean economy has been weakened by the rise of emerging economies. Furthermore, technological advantage of emerging economies has been strengthened. In this situation, Korea needs to make efforts to enhance global competitiveness. The growth of developing countries provides a new opportunities for Korea for export demand. However, this situation can be recognized as threats for Korea because Korea has to compete with those countries to expand market share. In this regard, reviewing the approaches of manufacturing innovation in German is important because German keeps remaining a high levels of competitiveness in spite of a rise of emerging economies and European recession. To do this, this research can give hints to advance the industrial policy improvements.