http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
다중노즐 3D프린터의 모니터링 시스템을 활용한 사용자 맞춤형 근골격계 재활기기 소재에 관한 연구
이선우(S. W. Lee),한현정(H. J. Han),최민석(M. S. Choi),이선곤(S. K. Lee),김주형(J. H. Kim),김용래(Y. R. Kim),권장우(J. W. Kwon) 한국재활복지공학회 2018 한국재활복지공학회 학술대회논문집 Vol.2018 No.4
In this paper we implement a remote monitoring system to create customized musculoskeletal rehabilitation devices, introduce the structures and properties of 3D printer filament materials which have the flexible tactile sensor entered into the rehabilitation equipment material with multiple nozzle 3D printers In addition, the material with the properties of the filament flexible tactile sensor proposed in this paper is a brittle and elastic material, and its length is adjusted by using a micro-stage within the elastic range limits, and the resistance value is measured and analyzed. As a result, the resistance increased up to 190%. As the experiment repeated, the conductive filament did not win the tension and was cut off, resulting in infinite divergence and disconnection of the electrode.
딥러닝 기반의 시각장애인 일상생활 보조용 임베디드 시스템 개발
양호준(H. J. Yang),장석우(S. W. Jang),박준수(J. S. Park),권장우(J. W. Kwon) 한국재활복지공학회 2021 한국재활복지공학회 학술대회논문집 Vol.2021 No.11
The purpose of this paper is to help improve the daily living conditions of blind people by integrating deep learning-based models capable of object detection, text detection and recognition, and cameras capable of depth measurement into one embedded system. Through the Mixed Precision method in the learning process, learning was made quickly on the Nvidia Jetson-affiliated board, and the Half Precision method was applied to increase the speed of inference in the actual environment.
Depth Camera를 이용한 관절 가동 범위 측정에 대한 연구
안영샘(Y. S. Ahn),설현우(H. W. Seol),주성수(S. S. Joo),김원진(W. J. Kim),박은식(E. S. Park),권장우(J. W. Kwon) 한국재활복지공학회 2017 한국재활복지공학회 학술대회논문집 Vol.2017 No.4
In this paper, we investigated the range of joint motion using Kinect, a Depth Camera. Today, as patients interest in medical information grows, there is a need to show patients quantified data. In the current hospital, a physician uses a goniometer to measure the range of motion of a patient. However, due to limited workforce and the status of the patient within a limited period of time, there is a limit to efficiently inspect and treat patients. In order to provide efficient treatment, this paper proposes a measurement method of the range of motion of the limb using the Depth Camera. We used the program to measure more quickly than the conventional measurement method. The experimental result shows that the error is within 3% of the actual measurement method. In addition, using the Depth Camera, it is possible to store and manage the range of motion of the joints as quantified data.
DBM을 활용한 EMR 기반의 스트로크환자의 재활프로토콜 추천 시스템
이선우(S. W. Lee),박경남(K. N. Park),김세진(S. J. Kim),원인수(I. S. Won),권장우(J. W. Kwon) 한국재활복지공학회 2017 한국재활복지공학회 학술대회논문집 Vol.2017 No.4
Most existing stroke rehabilitation training systems were determined by empirical treatment methods of doctors. Standard treatment guidelines for stroke rehabilitation have been introduced to address these problems. However, these guidelines are determined by the doctor"s objective judgment, and sometimes the doctor suggests different treatments for the same patient. In addition, there are cases where only the patient"s exercise function is presented and the individual"s risk management factor is missing during rehabilitation treatment. To solve these problems, this paper proposes a training method suitable for the patient using DBN (Deep Belief Network). In this method, a part of the patient"s EMR(Electronic Medical Records) data and a risk management factor are inputted and a method of presenting a suitable exercise to the patient is derived.
이경훈(K. H. Lee),길마로(M. R. Gil),손준열(J. Y. Sohn),정지욱(J. O. Chung),이광식(G. S. Lee),김규환(G. H. Kim),변기훈(K. H. Byun),권장우(J. W. Kwon) 한국재활복지공학회 2016 한국재활복지공학회 학술대회논문집 Vol.2016 No.11
In industrial site, there are many significant risks such as gas leak, falling and collision. Moreover, It could lead to a big accident. According to official statistics, gas explosions are about 100 to 200 cases in a year. However, worker can’t react to them. In this paper, we suggest a system that notify dangerous situation in a field and send warning signal to workers.