• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권4호
• /
• pp.1246-1262
• /
• 2021

Recent studies have demonstrated the strong ability of deep convolutional neural networks (CNNs) to significantly boost the performance in single image super-resolution (SISR). The key concern is how to efficiently recover and utilize diverse information frequencies across multiple network layers, which is crucial to satisfying super-resolution image reconstructions. Hence, previous work made great efforts to potently incorporate hierarchical frequencies through various sophisticated architectures. Nevertheless, economical SISR also requires a capable structure design to balance between restoration accuracy and computational complexity, which is still a challenge for existing techniques. In this paper, we tackle this problem by proposing a competent architecture called Enhanced U-Net Network (EUN), which can yield ready-to-use features in miscellaneous frequencies and combine them comprehensively. In particular, the proposed building block for EUN is enhanced from U-Net, which can extract abundant information via multiple skip concatenations. The network configuration allows the pipeline to propagate information from lower layers to higher ones. Meanwhile, the block itself is committed to growing quite deep in layers, which empowers different types of information to spring from a single block. Furthermore, due to its strong advantage in distilling effective information, promising results are guaranteed with comparatively fewer filters. Comprehensive experiments manifest our model can achieve favorable performance over that of state-of-the-art methods, especially in terms of computational efficiency.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권9호
• /
• pp.2882-2903
• /
• 2022

Image super-resolution (SR) processing is of great value in the fields of digital image processing, intelligent security, film and television production and so on. This paper proposed a densely connected deep learning network based on cascade architecture, which can be used to solve the problem of super-resolution in the field of image quality enhancement. We proposed a more efficient residual scaling dense block (RSDB) and the multi-channel cascade architecture to realize more efficient feature reuse. Also we proposed a hybrid loss function based on L₁ error and L_∞ error to achieve better L_∞ error performance. The experimental results show that the overall performance of the network is effectively improved on cascade architecture and residual scaling. Compared with the residual dense net (RDN), the PSNR / SSIM of the new method is improved by 2.24% / 1.44% respectively, and the L_∞ performance is improved by 3.64%. It shows that the cascade connection and residual scaling method can effectively realize feature reuse, improving the residual convergence speed and learning efficiency of our network. The L_∞ performance is improved by 11.09% with only a minimal loses of 1.14% / 0.60% on PSNR / SSIM performance after adopting the new loss function. That is to say, the L_∞ performance can be improved greatly on the new loss function with a minor loss of PSNR / SSIM performance, which is of great value in L_∞ error sensitive tasks.

• 한국의학물리학회지:의학물리
• /
• 제32권4호
• /
• pp.159-164
• /
• 2021

Purpose: We investigated the properties of CLEANBOLUS based on silicone with suitable characteristics for clinical use. Methods: We evaluated the characteristics of CLEANBOLUS and compared the results with the commercial product (Super-Flex bolus). Also, we conducted physical evaluations, including shore hardness, element composition, and elongation break. Transparency was investigated through the measured absorbance within the visible region (400-700 nm). Also, dosimetric characteristics were investigated with surface dose and beam quality. Finally, the volume of unwanted air gap was investigated based on computed tomography images for breast, chin, and nose using Super-Flex bolus and CELANBOLUS. Results: CLEANBOLUS showed excellent physical properties for a low shore hardness (000-35) and elongation break (>1,000%). Additionally, it was shown that CLEANBOLUS is more transparent than Super-Flex bolus. Dosimetric results obtained through measurement and calculation have an electron density similar to water in CLEANBOLUS. Finally, CLEANBOLUS showed that the volume of unwanted air gap between the phantom and each bolus is smaller than Super-Flex bolus for breast, chin, and nose. Conclusions: The physical properties of CLEANBOLUS, including excellent adhesive strength and lower shore hardness, reduce unwanted air gaps and ensure accurate dose distribution. Therefore, it would be an alternative to other boluses, thus improving clinical use efficiency.

• Carbon letters
• /
• 제12권4호
• /
• pp.243-248
• /
• 2011

Two different types of graphite, such as flake graphite (FG) and spherical graphite (SG), were used as anode materials for a lithium-ion secondary battery in order to investigate their electrochemical performance. The FG particles were prepared by pulverizing natural graphite with a planetary mill. The SG particles were treated by immersing them in acid solutions or mixing them with various carbon additives. With a longer milling time, the particle size of the FG decreased. Since smaller particles allow more exposure of the edge planes toward the electrolyte, it could be possible for the FG anodes with longer milling time to deliver high reversible capacity; however, their initial efficiency was found to have decreased. The initial efficiency of SG anodes with acid treatments was about 90%, showing an over 20% higher value than that of FG anodes. With acid treatment, the discharge rate capability and the initial efficiency improved slightly. The electrochemical properties of the SG anodes improved slightly with carbon additives such as acetylene black (AB), Super P, Ketjen black, and carbon nanotubes. Furthermore, the cyclability was much improved due to the effect of the conductive bridge made by carbon additives such as AB and Super P.

• 한국항만경제학회지
• /
• 제26권3호
• /
• pp.18-41
• /
• 2010

본 논문에서는 Liu(2009)의 슬랙변수모형과 윌콕슨 부호순위 검정방법을 도입하고 새롭게 수퍼효율성모형을 도입한 Tone(2002)의 수퍼SBM모형에 대해서 이론적으로 설명하였다. 또한 1997년부터 2007년까지 국내 20개 항만의 5개의 산출물(항만서비스만족도, 수출입물량, 선박입출항척수, 항만재정수입, 컨테이너화물처리량)과 3개의 투입물(항만투자금액, 접안능력, 하역능력)을 이용하여 국내항만들의 효율성을 측정한 후에, 효율성 순위를 이용하여 윌콕슨의 부호순위검정을 통해서 수퍼SBM모형이 갖고 있는 항만효율성측면의 예측력을 측정하는 방법을 보여 주었다. 또한 개별항만별, 연도별로 적정한 항만투자의 금액을 예측하여 보여 주었다. 실증분석의 주요한 결과는 다음과 같다. 첫째, Wilcoxon의 부호순위 검정통계량에 의거하여 효율성 예측에 대한 가설검정을 해 보면, 제1모형에서는 p값이 평균 97.68% 수준에서 성과예측력을 가지는 것으로 나타났으며, 제2모형에서는 약 76.1%로 나타났다. 둘째, 수퍼SBM모형에 의한 항만별 항만투자의 적정투자규모를 정확하게 제시하였다. 셋, 제2모형에 대한 수퍼SBM모형에 의한 항만별 항만투자의 적정투자규모예측 결과는 다음과 같다. (ㄱ) 항만투자규모가 적정했던 항만들은 인천, 완도, 여수, 서귀포, 부산, 울산, 속초로 나타났다. (ㄴ) 거의 모든 항만들에서 투자금액이 과다한 것으로 나타났으며, 과소한 항만들은 삼천포, 마산, 울산, 삼척, 동해항으로 나타났다. (ㄷ) 차기년도의 항만의 적정투자규모를 예측하여 제시할 수는 있었다. 본 연구의 정책적인 함의는 국내항만관리자들은 항만의 환경이 변하고 있으므로 본 연구에서 제시하고 있는 수퍼SBM모형을 이용하여 항만들의 항만투자의 효율성 측정이나, 경영성과를 예측하는데 도입을 해야만 한다.

• Journal of Biosystems Engineering
• /
• 제43권2호
• /
• pp.94-102
• /
• 2018

Purpose: In this study, a maintenance free super-capacitor battery charging system based on the photovoltaic module, to be used in agricultural electric carriers, was developed and its charging characteristics were studied in detail. Methods: At first, the electric carrier system configuration is introduced and the electric control components are presented. The super-capacitor batteries and photovoltaic module used in the experiment are specified. Next, the developed charging system consisting of a constant current / constant voltage Buck converter as the charging device and a super-capacitor cell as a balancing device are initiated. The proposed circuit design, a developed PCB layout of each device and a proportional control to check the current and voltage during the charging process are outlined. An experiment was carried out using a developed prototype to clarify the effectiveness of the proposed system. A power analyzer was used to measure the current and voltage during charging to evaluate the efficiency of the energy storage device. Finally, the conclusions of this research are presented. Results: The experimental results show that the proposed system successfully controls the charging current and balances the battery voltage. The maximum voltage of the super-capacitor battery obtained by using the proposed battery charger is 16.2 V, and the maximum charging current is 20 A. It was found that the charging time was less than an hour through the duty ratio of 95% or more. Conclusions: The developed battery charging system was successfully implemented on the agricultural electric carriers.

• 한국건축시공학회:학술대회논문집
• /
• 한국건축시공학회 2008년도 추계 학술논문 발표대회
• /
• pp.37-41
• /
• 2008

According to the recent rapidly increasing that construction works are gradually Manhattanized mainly the grand scaled residential buildings, the foundation of the building that is related to safety is increasing for building as a grand scaled mat concrete. Because mat concrete can not be simultaneously placing of concrete in a great quantity due to the circumstance at the field, the inequal deformation of the tensile stress that according to the time lag of hydration heat between the upper layer and the lower layer is affecting as a cause that is the possibility of crack occurrence by increasing. Accordingly, this research checked the efficiency of super retard concrete in applying real structures, and we implemented the preparatory experiment to settle up the inequal deformation of the tensile stress substantially that is according to the time lag of placement between the upper layer and the lower layer by controlling the setting time using the super retarding agent. As the result of test, the more target of delay time lengthened, the more fluidity increased and air content indicated a little differences. There was from 2 to 10 hours between the standard curing and the outside curing at the setting time and in case of calculating the rate of mixing at real structure is required that mix promotion, increasing the amount of mixing, by setting up the curing temperature. The super retard concrete showed the result that in compressive strength, early-age strength was smaller than normal concrete whereas it was same or more figures from at the aging 28days because of the super retarding agent.

• 산업경영시스템학회지
• /
• 제40권4호
• /
• pp.21-28
• /
• 2017

Improving efficiency of the telecommunication is crucial to the development and growth of Korean economy. Recently, it has become important with the huge development of information technology and its greater potential for extensive impact on the rest of the economy. Hence, it is useful to determine the factors that help enhance efficiency in telecommunication and consider them in improving the evaluation model. This study applies DEA (data envelopment analysis) to evaluate the relative efficiency of 51 branches of a Korean telecommunication company. Using the super-efficiency approach, we tested outliers which may affect the results and ranked the efficient branches. A method of deriving key variables applied to business operation is proposed to identify the key performance indicators for evaluation that takes environmental (non-discretionary) factors into account. We used the extended CCR model proposed by Banker and Morey to investigate the influence of non-discretionary factor. The information provided by the model (slacks, weights) and the sensitivity analysis shows that the most important indicator that affects the branch performance is operating cost. The results of sensitivity analysis show that average efficient score decreases from 0.972 (base case) to 0.863 for CASE2-COST. The average score of the data proves the priority of operating cost over other indicators. The effect of environmental (non-discretionary) variable was found to be significant. The population effect was positive and improved overall efficiency by 0.91% on average. Non-discretionary factor plays a meaningful role explaining the performance of branches. The performance optimization report can help a manager of an inefficient branch to develop branch strategies. Managers can identify the top-performing units, study best practices and adopt the strategy to the organization.

• 서비스연구
• /
• 제11권4호
• /
• pp.43-55
• /
• 2021

본 연구는 OECD 회원국들을 대상으로 의료서비스의 운영효율성과 품질효율성을 비교 분석하였다. 이를 위하여 2017~2019년의 OECD 보건통계(Health Statistics) 자료를 이용하였다. 분석 방법으로는 산출지향형(output-oriented) 규모수익가변(VRS: Variable Returns to Scale) 모형을 적용하여 초효율성(super efficiency)을 측정하였다. 분석결과, 운영효율성의 상위 그룹에는 스위스, 한국, 이탈리아, 노르웨이, 오스트리아, 네덜란드, 중위 그룹에는 캐나다, 그리스, 덴마크, 멕시코, 미국, 영국, 체코, 프랑스, 핀란드, 헝가리, 호주, 하위 그룹에는 벨기에, 독일, 스페인이 포함되었다. 품질효율성의 상위 그룹에는 노르웨이, 스위스, 스페인, 중위 그룹에는 그리스, 덴마크, 멕시코, 미국, 영국, 캐나다, 프랑스, 핀란드, 헝가리, 호주, 하위 그룹에는 네덜란드, 독일, 벨기에, 오스트리아, 이탈리아, 체코, 한국이 해당되었다. 2018년을 기준으로 OECD 회원국별로 효율성을 비교 분석한 결과, 운영효율성이 가장 효율적인 한국은 품질효율성이 비효율적이었다. 한국(0.998)은 그리스(0.422), 스위스(0.207), 스페인(0.371)을 벤치마킹하여 기대수명을 0.2(0.2%), 주관적 건강 인지를 44.2(138.1%)씩 개선하여 품질효율성을 향상시켜야 한다. 이는 국내에서 기대수명이나 유병률 같은 객관적 지표로 건강상태를 측정하고 있지만, 의학적 진단 여부와 별개로 국민들이 본인의 건강에 관하여 주관적으로 평가하는 것을 관리할 필요가 있음을 의미한다. 본 연구는 운영효율성에 초점을 맞춘 대부분의 선행연구들과 다르게 운영효율성과 품질효율성을 함께 측정했다는 점에서 학술적인 시사점을 지닌다. 또한 OECD 회원국별 의료서비스 산업의 효율성을 비교 분석함으로써, 국내 의료서비스 산업의 국제적인 경쟁력 수준을 확인하고 효율성 개선 방안을 제시했다는 점에서 실무적인 시사점을 지닌다.

• 대한전자공학회논문지SP
• /
• 제48권5호
• /
• pp.129-136
• /
• 2011

본 논문에서는 입력신호에 따라 효율적인 부호화를 위해 다중모드를 가지는 초광대역 부호화기를 제안하였다. 최근에 개발된 G.718기반의 초광대역 확장 코덱은 초광대역에 존재하는 오디오 신호를 부호화하기 위하여 Generic모드와 Sinusoidal모드로 신호를 분류하여 부호화 하지만, 현악기, 관현악기와 같은 오디오 신호에 존재하는 멀티플 피치 및 하모닉 성분과 타악기와 같은 오디오 신호에 존재하는 Individual-Line 성분을 효율적으로 부호화하지 못하였다. 제안하는 방법은 오디오 신호에 존재하는 하모닉과 Individual-Line성분의 특징을 고려하여 모드를 세분화하였다. 성능평가를 위하여 객관적인 평가로 MDCT 도메인에서 SNR을 이용하였으며 MUSHRA 테스트를 통해 주관적인 평가를 하였다. 실험 결과 제안한 방법은 기존의 G.718 초광대역 부호화기보다 Individual-Line신호 경우 약 3dB, 하모닉 신호의 경우 약 0.8dB 높은 SNR을 보였으며 MUSHRA 테스트 역시 평균 5점 음질 향상을 확인하였다.