• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2011.11a
• /
• pp.143-146
• /
• 2011

This paper presents human perception-based video coding method using an online learning framework. In this work, we analyze the relationship between human attention regions and video quality, and also consider human memory. We classify the motion patterns based on the analysis. Then, we devise a motion pattern classification method using Hedge algorithm. Along with the motion patterns, we smooth out the specific regions or sharpen details of the regions using the regional priorities. The preprocessed sequences are applied to the video codec. The performance is excellent on the overall quality as well as the regional quality.

• Food Science and Biotechnology
• /
• v.18 no.5
• /
• pp.1243-1252
• /
• 2009

User-friendly software was developed to determine the shelf-life of perishable Korean seasoned side dishes in real time based on growth models of spoilage and pathogenic microorganisms. In the program algorithm, the primary spoilage and fastest-growing pathogenic organisms are selected according to the product characteristics, and their growth is simulated based on the previously monitored or recorded temperature history. To predict the growth of spoilage organisms with confidence limits, kinetic models for aerobic bacteria or molds/yeasts from published works are used. Growth models of pathogenic bacteria were obtained from the literature or derived with regression of their growth rate data estimated from established software packages. These models are also used to check whether the risk of pathogenic bacterial growth exceeds that of food spoilage organisms. Many example simulations showed that the shelf-lives of the examined foods are predominantly limited by the growth of spoilage organism rather than by pathogenic bacterial growth.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.5
• /
• pp.2590-2606
• /
• 2017

Generally, research of classical image classification algorithms assume that training data and testing data are derived from the same domain with the same distribution. Unfortunately, in practical applications, this assumption is rarely met. Aiming at the problem, a domain adaption image classification approach based on multi-sparse representation is proposed in this paper. The existences of intermediate domains are hypothesized between the source and target domains. And each intermediate subspace is modeled through online dictionary learning with target data updating. On the one hand, the reconstruction error of the target data is guaranteed, on the other, the transition from the source domain to the target domain is as smooth as possible. An augmented feature representation produced by invariant sparse codes across the source, intermediate and target domain dictionaries is employed for across domain recognition. Experimental results verify the effectiveness of the proposed algorithm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.8
• /
• pp.4090-4102
• /
• 2018

After emerging online communications, text mining and sentiment analysis has been frequently applied into analyzing electronic word-of-mouth. This study aims to develop a domain-specific lexicon of sentiment analysis to predict box office success in Korea film market and validate the feasibility of the lexicon. Natural language processing, a machine learning algorithm, and a lexicon-based sentiment classification method are employed. To create a movie domain sentiment lexicon, 233,631 reviews of 147 movies with popularity ratings is collected by a XML crawling package in R program. We accomplished 81.69% accuracy in sentiment classification by the Korean sentiment dictionary including 706 negative words and 617 positive words. The result showed a stronger positive relationship with box office success and consumers' sentiment as well as a significant positive effect in the linear regression for the predicting model. In addition, it reveals emotion in the user-generated content can be a more accurate clue to predict business success.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.21 no.6
• /
• pp.1379-1389
• /
• 1996

Hangul character always consists of consonants-vowel-consonants in order. Using this point, this paper proposes an approach to design a model for spotting each letter in Hangul, and then recognize characters based on the spotting results. The network model consist of a set of HMMs. The letter search is carried out by Viterbi algorithm, while character recognition is performed by searching the lattice of letter hypotheses. Experimental results show that, in spite of simple architecture of recognition, the performance is quite high reaching 87.47% for discrete regular characters. In particular the approach shows highly plausible segmentation of letters in characters.

• International Journal of Control, Automation, and Systems
• /
• v.6 no.4
• /
• pp.596-606
• /
• 2008

In this paper, we present a co-design methodology of dynamic optimal network-bandwidth allocation (ONBA) and adaptive control for networked control systems (NCSs) to optimize overall control performance and reduce total network-bandwidth usage. The proposed dynamic co-design strategy integrates adaptive feedback control with real-time scheduling. As part of this co-design methodology, a "closed-loop" ONBA algorithm for NCSs with communication constraints is presented. Network-bandwidth is dynamically assigned to each control loop according to the quality of performance (QoP) information of each control loop. As another part of the co-design methodology, a network quality of service (QoS)-adaptive control design approach is also presented. The idea is based on calculating new control values with reference to the network QoS parameters such as time delays and packet losses measured online. Simulation results show that this co-design approach significantly improves overall control performance and utilizes less bandwidth compared to static strategies.

• Proceedings of the KIPE Conference
• /
• 2018.07a
• /
• pp.240-242
• /
• 2018

본 논문은 배터리 등가 회로 모델의 파라미터를 실시간으로 추정 가능한 온라인 업데이트 알고리즘을 비교 분석한다. 온라인 업데이트 알고리즘은 EKF (Extended Kalman Filter)와 RLS (Recursive Least Square)를 기반으로 구성되며 Li-ion 배터리를 이용한 실험 결과를 통하여 각 알고리즘의 연산량, 정확도 등을 분석한다. 분석된 결과를 바탕으로 적용 시스템의 환경과 특성에 따라 적합한 온라인 업데이트 알고리즘을 제안한다.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2018.11a
• /
• pp.256-257
• /
• 2018

This paper deals with the path-following control for marine surface vehicles with underactuated characteristics. In consideration of practical limitations of actuators, an improved DVS(dynamic virtual ship) guidance algorithm is proposed with the multi-level DVS optionally selected to be tracked. To address the output-feedback control issue, an adaptive FLS(fuzzy logical systems) is devised to online approximate the kinematic states. Based on that observing framework, the path-following control law is thereafter derived. Simulations testify effectiveness of the proposed scheme

• Journal of the Optical Society of Korea
• /
• v.20 no.6
• /
• pp.770-777
• /
• 2016

It has been a challenge for researchers to accurately measure high temperature creep strain online without damaging the mechanical properties of the pipe surface. To this end, a noncontact method for measuring high temperature strain of a main steam pipe based on digital image correlation was proposed, and a system for monitoring of high temperature strain was designed and developed. Wavelet thresholding was used for denoising measurement data. The sub-pixel displacement search algorithm with curved surface fitting was improved to increase measurement accuracy. A field test was carried out to investigate the designed monitoring system of high temperature strain. The measuring error was less than $0.4ppm/^{\circ}C$, which meets actual measurement requirements for engineering. Our findings provide a new way to monitor creep damage of the main steam pipe of a boiler of an ultra-supercritical power plant in service.

• Proceedings of the KSR Conference
• /
• 2006.11b
• /
• pp.742-747
• /
• 2006

This paper describes online stator-resistance estimation of a linear induction motor(LIM) with cage-type secondary using direct thrust control(DTC), where the resistance value is derived from stator-winding temperature estimation using thermocouple. In this paper, corrected stator resistance has an error in actuality measurement resistance. So compensation coefficient $\kappa$ which is decided through comparison and verifying several times relation of calculated resistance and measured motor line-line resistance. The stator-winding temperature information can also be used for monitoring, protection, and fault-tolerant control of the machine. Also, this paper reports the LIM's responses of the flux measured by the proposed stator resistance compensation algorithm.