http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
진용식,이민호,Jin, Yongsik,Lee, Minho 대한임베디드공학회 2017 대한임베디드공학회논문지 Vol.12 No.1
We propose a face recognition and notification system that can transform visual face information into tactile signals in order to help visually impaired people. The proposed system consists of a glasses type camera, a mobile computer and an electronic cane. The glasses type camera captures the frontal view of the user, and sends this image to mobile computer. The mobile computer starts to search for human's face in the image when obstacles are detected by ultrasonic sensors. In a case that human's face is detected, the mobile computer identifies detected face. At this time, Adaboost and compressive sensing are used as a detector and a classifier, respectively. After the identification procedures of the detected face, the identified face information is sent to controller attached to a cane using a Bluetooth communication. The controller generates motor control signals using Pulse Width Modulation (PWM) according to the recognized face labels. The vibration motor generates vibration patterns to inform the visually impaired person of the face recognition result. The experimental results of face recognition and notification system show that proposed system is helpful for visually impaired people by providing person identification results in front of him/her.
휴머노이드 로봇의 거리 추정능력과 행동 개선을 위한 자율학습 알고리즘
진용식(Yongsik Jin),이민호(Minho Lee) 대한전자공학회 2016 대한전자공학회 학술대회 Vol.2016 No.11
In this paper, we propose a new action and perception cycle learning framework for improving action and perception simultaneously. In the leaning of depth estimation, the sensory invariance driven action is used to provide useful sensory data to a robot. At this time, object size information is used as a clue of target distance, and gives suitable timing for depth estimation learning. By using improved depth estimation, the robot can perceive current spatial state in surrounding environment more accurately. Thus, the action generation of robot can be improved. The proposed learning framework is implemented in humanoid robot (Nao) which is additionally mounted two cameras. Experimental results shows that the proposed learning framework can improve depth estimation and action generation performance simultaneously.
[자율이동체 정보처리] 다중 센서를 가지는 상태추정기 기반의 자율 이동로봇 추적제어 시스템 구현
진용식(Yongsik Jin),한승용(Seungyong Han),이상문(Sangmoon Lee) 대한전기학회 2019 전기학회논문지 Vol.68 No.5
In this paper, we propose a tracking control method for autonomous mobile robot based on state estimation with multi-sensor. The proposed system consists of a camera and encoder to measure the accurate position and orientation of a robot. In order to fuse data obtained from different sensors, we designed a state estimator with multi-rate sampled-data. The performance of the tracking control is guaranteed by the explicit model predictive control method which is employed for real-time implementation. For autonomous navigation, we implemented camera based sign recognition, obstacle location determination, and path generation algorithm using rider data. The proposed autonomous tracking control system for mobile robot is developed in the embedded environment based on the robot operating system (ROS). Virtual simulations and real experiments are performed to show the effectiveness and validity of the proposed method.
권우경,진용식,이상준 대한임베디드공학회 2021 대한임베디드공학회논문지 Vol.16 No.2
Human-robot interaction has received a lot of attention as collaborative robots became widely used in many industrial applications. This paper proposes a deep learning method for collision identification of collaborative robots. This method expands the idea of CollisionNet, which was proposed for collision detection, to identify locations of collisions. Collision identification is far more difficult compared to collision detection, because sensor data are highly correlated when collisions occur at close locations. To improve the identification accuracy, this paper proposes an auxiliary loss, which is called consistency loss. This auxiliary loss guides the training of a deep neural network to predict consistent predictions for each single collision event. In experiments, we demonstrate the effectiveness of the proposed method.
수열법으로 성장한 ZnO Nanorod/ZnO/Si(100)의 특성
정민호,진용식,최성민,한덕동,최대규,Jeong, Min-Ho,Jin, Yong-Sik,Choi, Sung-Min,Han, Duk-Dong,Choi, Dae-Kue 한국재료학회 2012 한국재료학회지 Vol.22 No.4
Nanostructures of ZnO, such as nanowires, nanorods, nanorings, and nanobelts have been actively studied and applied in electronic or optical devices owing to the increased surface to volume ratio and quantum confinement that they provide. ZnO seed layer (about 40 nm thick) was deposited on Si(100) substrate by RF magnetron sputtering with power of 60 W for 5 min. ZnO nanorods were grown on ZnO seed layer/Si(100) substrate at $95^{\circ}C$ for 5 hr by hydrothermal method with concentrations of $Zn(NO_3)_2{\cdot}6H_2O$ [ZNH] and $(CH_2)_6N_4$ [HMT] precursors ranging from 0.02M to 0.1M. We observed the microstructure, crystal structure, and photoluminescence of the nanorods. The ZnO nanorods grew with hexahedron shape to the c-axis at (002), and increased their diameter and length with the increase of precursor concentration. In 0.06 M and 0.08 M precursors, the mean aspect ratio values of ZnO nanorods were 6.8 and 6.5; also, ZnO nanorods had good crystal quality. Near band edge emission (NBE) and a deep level emission (DLE) were observed in all ZnO nanorod samples. The highest peak of NBE and the lower DLE appeared in 0.06 M precursor; however, the highest peak of DLE and the lower peak of NBE appeared in the 0.02 M precursor. It is possible to explain these phenomena as results of the better crystal quality and homogeneous shape of the nanorods in the precursor solution of 0.06 M, and as resulting from the bed crystal quality and the formation of Zn vacancies in the nanorods due to the lack of $Zn^{++}$ in the 0.02 M precursor.
CPPS를 위한 산업용 매니플레이터의 힘 센서리스 외력 추정기 기반 적응 임피던스 제어
박종천,한승용,진용식,이상문 대한임베디드공학회 2019 대한임베디드공학회논문지 Vol.14 No.5
This paper proposes a structure of a variable impedance control system based on sensor-less external force estimator of industrial manipulators for cyber physical production systems (CPPS). To implement CPPS, a feedback system is constructed by using the robot operating system (ROS) and an external force estimator which is designed to measure the external force applied to the manipulator without a force sensor. Based on the robot dynamics, the robot-human cooperating system for the cyber physics production system is implemented through a controller that changes the impedance characteristics of the manipulator according to the situation using the external force estimator. Simulation and experimental results verify the effectiveness of the proposed control system.