http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
다기능 안전고압호스를 이용한 하수관 비굴착 보수공법의 성능개선
최장환(Jang Hwan Choi),임봉수(Bong Su Lim),이원용(Won Yong Lee) 대전대학교 환경문제연구소 2016 환경문제연구소 논문집 Vol.20 No.-
This study was carried out to improve the performance of sewer trenchless entire repair method by using the multifunctional safety high-pressure hose. Because the flexural strength and flexing modulus of elasticity increased in the range from 60℃ to 80℃ of the setting temperature, the unform quality was secured to satisfy the standard quality and consolidate the strength, by supplying steam to the entire sewer through the safety high-pressure hose simultaneously. After the steam was supplied, the temperature of the condensate water with a large amount occurred , compared with a small amount, was very low. So, the setting time was shorter over than one hour by reducing the unsatisfied cure state and excluding the condensate water. The standard quality increased by approximately 15% at upper, central, and lower points of the same sewer was gained from the test result of CIPP(cured in place pipes) constructed at the optimum setting temperature. The steam and air were supplied evenly to the sewer by controlling the safety high-pressure hose in the tense and relaxed state and by using the exclusive function of condensate water. The availability and safety for the constructed process were achieved.
최장환(Jang-Hwan Choi),강지원(Ji-Won Kang),김동욱(Dong-Wook Kim),서영호(Young_Ho Seo) 한국정보통신학회 2021 한국정보통신학회 종합학술대회 논문집 Vol.25 No.1
본 논문에서는 딥러닝 네트워크를 이용한 고해상도 프린지 패턴 생성 기법을 제안한다. 컴퓨터를 이용하여 홀로그램을 생성하기 위해서는 매우 방대한 계산이 필요하다. 이를 대체할 수단으로 딥러닝을 채택하여 대체 가능함을 보였으나 출력되는 프린지 패턴 해상도의 한계가 존재하였다. 이를 개선하기 위한 고해상도 프린지 패턴 생성을 위한 기법을 제안한다. In this paper, we propose a high-resolution fringe pattern generation technique using deep learning networks. Generating a hologram using a computer requires a very large amount of computation. Therefore, in order to replace this, it was shown that it can be replaced through deep learning, but there was a limitation in the resolution of the output fringe pattern. To improve this, we propose an algorithm for generating a high-resolution fringe pattern.
엣지 가이드 복합 손실을 사용한 다중 프레임 기반 LDCT 화질 개선 프레임워크
전선영(Sun-Young Jun),최장환(Jang-Hwan Choi) 대한전자공학회 2024 대한전자공학회 학술대회 Vol.2024 No.6
Low-dose computed tomography (LDCT) has become an essential tool for reducing radiation exposure while still providing vital anatomical information necessary for diagnosing various pathologies. However, the inevitable loss of information, including issues like noise, streak artifacts, and blurred details, can adversely affect clinical diagnostics. To enhance LDCT imaging, there is notable potential in effectively minimizing image noise while retaining crucial anatomical features. This paper introduces an innovative dual-branch network architecture for LDCT denoising that combines edge-guided compound loss with multiple optimal image-based techniques. We evaluate the denoising performance by comparing various the latest state-of-the-art (SOTA) algorithms and demonstrating that our proposed model surpasses the SOTA algorithms. The proposed method shows that both the objective and perceptual quality of LDCT images are improved.
딥러닝 표정 인식을 활용한 실시간 온라인 강의 이해도 분석
이자연,정소현,신유원,이은혜,하유빈,최장환,Lee, Jaayeon,Jeong, Sohyun,Shin, You Won,Lee, Eunhye,Ha, Yubin,Choi, Jang-Hwan 한국멀티미디어학회 2020 멀티미디어학회논문지 Vol.23 No.12
Due to the spread of COVID-19, the online lecture has become more prevalent. However, it was found that a lot of students and professors are experiencing lack of communication. This study is therefore designed to improve interactive communication between professors and students in real-time online lectures. To do so, we explore deep learning approaches for automatic recognition of students' facial expressions and classification of their understanding into 3 classes (Understand / Neutral / Not Understand). We use 'BlazeFace' model for face detection and 'ResNet-GRU' model for facial expression recognition (FER). We name this entire process 'Degree of Understanding (DoU)' algorithm. DoU algorithm can analyze a multitude of students collectively and present the result in visualized statistics. To our knowledge, this study has great significance in that this is the first study offers the statistics of understanding in lectures using FER. As a result, the algorithm achieved rapid speed of 0.098sec/frame with high accuracy of 94.3% in CPU environment, demonstrating the potential to be applied to real-time online lectures. DoU Algorithm can be extended to various fields where facial expressions play important roles in communications such as interactions with hearing impaired people.