• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제8권5호
• /
• pp.205-212
• /
• 2019

장면 그래프는 영상 내 물체들과 각 물체 간의 관계를 나타내는 지식 그래프를 의미한다. 본 논문에서는 3차원 실내 환경을 위한 3차원 장면 그래프를 생성하는 모델을 제안한다. 3차원 장면 그래프는 물체들의 종류와 위치, 그리고 속성들뿐만 아니라, 물체들 간의 3차원 공간 관계들도 포함한다. 따라서 3차원 장면 그래프는 에이전트가 활동할 실내 환경을 묘사하는 하나의 사전 지식 베이스로 볼 수 있다. 이러한 3차원 장면 그래프는 영상 기반의 질문과 응답, 서비스 로봇 등과 같은 다양한 분야에서 유용하게 활용될 수 있다. 본 논문에서 제안하는 3차원 장면 그래프 생성 모델은 크게 물체 탐지 네트워크(ObjNet), 속성 예측 네트워크(AttNet), 변환 네트워크(TransNet), 관계 예측 네트워크(RelNet) 등 총 4가지 부분 네트워크들로 구성된다. AI2-THOR가 제공하는 3차원 실내 가상환경들을 이용한 다양한 실험들을 통해, 본 논문에서 제안한 모델의 높은 성능을 확인할 수 있었다.

• 대한기계학회논문집A
• /
• 제21권1호
• /
• pp.62-72
• /
• 1997

High flexibility and productivity using industrial robots are being achieved in manufacturing lines with off-line robot programmings. A good off-line programming system should have functions of robot modelling, trajectory planning, graphical teach-in, kinematic and dynamic simulations. Simulated results, however, can hardly be applied to on-line tasks until any calibration procedure is accompained. This paper proposes a visual calibration scheme in order to provide a calibration tool for our own off-line programming system of SCARA robots. The suggested scheme is based on the position-based visual servoings, and the perspective projection. The scheme requires only one camera as it uses saved kinematic data for three-dimensional visual calibration. Predicted images are generated and then compared with camera images for updating positions and orientations of objects. The scheme is simple and effective enough to be used in real time robot programming.

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 2002년도 춘계학술대회 논문집
• /
• pp.567-572
• /
• 2002

The equipment of industrial robot in manufacturing and assembly lines has rapidly increased. In order to achieve high productivity and flexibility, it becomes very important to develop the visual feedback control system with Off-Line Programming System(OLPS ). We can save much efforts and time in adjusting robots to newly defined workcells by using Off-Line Programming System. A proposed visual calibration scheme is based on position-based visual feedback. The visual calibration system is composed of a personal computer, an image processing board, a video monitor, and one camera. The calibration program firstly generates predicted images of objects in an assumed end-effector position. The process to generate predicted images consists of projection to screen-coordinates, visible range test, and construction of simple silhouette figures. Then, camera images acquired are compared with predicted ones for updating position and orientation data. Computation of error is very simple because the scheme is based on perspective projection, which can be also expanded to experimental results. Computation time can be extremely reduced because the proposed method does not require the precise calculation of tree-dimensional object data and image Jacobian.

• 에너지공학
• /
• 제29권3호
• /
• pp.50-56
• /
• 2020

Currently, 3D printing technology is a new revolutionary additive manufacturing process that can be used for making three dimensional solid objects from digital films. In 2019, this 3D printing technology spreading vigorously in production parts (57%), bridge production (39%), tooling, fixtures, jigs (37%), repair, and maintenance (38%). The applications of 3D printing are expanding to the defense, aerospace, medical field, and automobile industry. The raw materials are playing a key role in 3D printing. Various additive materials such as plastics, polymers, resins, steel, and metals are used for 3D printing to create a variety of designs. The main advantage of the green cement for 3D printing is to enhance the mechanical properties, and durability to meet the high-quality material using in construction. There are several advantages with 3D printing is a limited waste generation, eco-friendly process, economy, 20 times faster, and less time-consuming. This research article reveals that the role of green cement as an additive material for 3D printing.

• 한국군사과학기술학회지
• /
• 제16권5호
• /
• pp.653-658
• /
• 2013

For towing the new type armored vehicle and maintaining the close support, the armored recovery vehicle(ARV) with winch and crane has been developed. In case of crane, it is mainly used to salvage heavy objects by rotational and vertical motion. Especially, the crane bracket is very important parts due to fixing the ARV's body and rotary joint and preventing the force rotation of crane. Therefore, the crane bracket needs to have an enough strength to endure the high load and it is very important to analyze the stress distribution under loads. In the present work, the experimental and analytical investigation on structural integrity evaluation of crane bracket were carried out. The simulation of three-dimensional finite element method(FEM) was compared with experimental datum. From the numerical results, the FEM simulations corresponded well with th experimental results and the structural safety was confirmed by safety factor.

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 2001년도 추계학술대회(한국공작기계학회)
• /
• pp.140-145
• /
• 2001

The equipment of industrial robot in manufacturing and assembly lines has rapidly increased. In order to achieve high productivity and flexibility, it becomes very important to develop the visual feedback control system with Off-Line Programming System(OLPS). We can save much efforts and time in adjusting robots to newly defined workcells by using Off-Line Programming System. A proposed visual calibration scheme is based on position-based visual feedback. The visual calibration system is composed of a personal computer, an image processing board, a video monitor, and one camera. The calibration program firstly generates predicted images of objects in an assumed end-effector position. The process to generate predicted images consists of projection to screen-coordinates, visible range test, and construction of simple silhouette figures. Then, camera images acquired are compared with predicted ones for updating position and orientation data. Computation of error is very simple because the scheme is based on perspective projection, which can be also expanded to experimental results. Computation time can be extremely reduced because the proposed method does not require the precise calculation of tree-dimensional object data and image Jacobian.

• 한국산업융합학회 논문집
• /
• 제17권2호
• /
• pp.54-60
• /
• 2014

The equipment of SCARA robot in processing and assembly lines has rapidly increased. In order to achieve high productivity and flexibility, it becomes very important to develop the visual feedback control system with Off-Line Programming System(OLPS). We can save much efforts and time in adjusting robots to newly defined workcells by using OLPS. A proposed visual calibration scheme is based on position-based visual feedback. The calibration program firstly generates predicted images of objects in an assumed end-effector position. The process to generate predicted images consists of projection to screen-coordinates, visible range test, and construction of simple silhouette figures. Then, camera images acquired are compared with predicted ones for updating position and orientation data. Computation of error is very simple because the scheme is based on perspective projection, which can be also expanded to experimental results. Computation time can be extremely reduced because the proposed method does not requirethe precise calculation of tree-dimensional object data and image Jacobian.

• 한국멀티미디어학회논문지
• /
• 제8권6호
• /
• pp.768-775
• /
• 2005

For sign languages are main communication means among hearing-impaired people, there are communication difficulties between speaking-oriented people and sign-language-oriented people. Automated sign-language recognition may resolve these communication problems. In sign languages, finger spelling is used to spell names and words that are not listed in the dictionary. There have been research activities for gesture and posture recognition using glove-based devices. However, these devices are often expensive, cumbersome, and inadequate for recognizing elaborate finger spelling. Use of colored patches or gloves also cause uneasiness. In this paper, a vision-based finger spelling recognition system is introduced. In our method, captured hand region images were separated from the background using a skin detection algorithm assuming that there are no skin-colored objects in the background. Then, hand postures were recognized using a two-dimensional grid analysis method. Our recognition system is not sensitive to the size or the rotation of the input posture images. By optimizing the weights of the posture features using a genetic algorithm, our system achieved high accuracy that matches other systems using devices or colored gloves. We applied our posture recognition system for detecting Korean Sign Language, achieving better than $93\%$ accuracy.

• 대한조선학회논문집
• /
• 제50권3호
• /
• pp.160-166
• /
• 2013

Recently submerged objects moving at high speed such as a supercavitating torpedo have been studied for their practical advantage of the dramatic drag reduction. In this study, we are focusing our attention on supercavitating flows around axisymmetric cavitators; a numerical method based on inviscid flow is developed and predicted supercavities around several shapes of 2D and 3D cavitators are presented. The results are validated by comparison of existing theoretical and empirical results. In addition, characteristic features of supercavity shapes and drag forces acting on a real scale torpedo are evaluated according to practically appropriate operating conditions.

• Journal of the Optical Society of Korea
• /
• 제13권3호
• /
• pp.379-385
• /
• 2009

The vergence control of binocular stereoscopic camera is the most essential factor for acquiring high quality stereoscopic images. In this paper, we proposed a binocular stereoscopic camera vergence control method using disparity information by the simple image processing and estimate the quantity of vergence control using the Lagrange interpolation equation. The method of extracting disparity information through image processing is as follows: first the key-object in left & right images was extracted through labeling of the central area of the image, and then a simple method was used for calculating the disparity value of the same key-object in the labeled left and right images. The vergence control method uses disparity information and keeps the convergence distance of left & right cameras and the distance of the key-object the same. According to the proposed method, variance in the distance of the key-object and application of calculated disparity information of obtained left & right images to the quadratic Lagrange interpolation equation could estimate the quantity of vergence control, which confirmed that the method of stereoscopic camera vergence control can be simplified through experiments on various key-objects and other convergence distance.