• 한국게임학회 논문지
• /
• 제10권5호
• /
• pp.103-113
• /
• 2010

본 논문은 얼굴의 표정과 몸 동작을 광학식 동작 포착장비를 활용하여 동시에 포착하는 경우에 있어 얼굴 부위 마커들에 대한 노이즈에 강건한 데이터 처리 방법에 대해 다룬다. 일반적인 얼굴 표정만 포착하는 경우와 달리, 몸의 움직임과 동시에 포착할 경우 포착용 카메라가 멀리 있어 얼굴에 붙인 마커들의 궤적 데이터는 특별한 처리를 요한다. 특히 궤적의 표식화, 빈 곳 메우기, 노이즈 제거의 과정이 필수적이며, 이러한 과정을 위해 본 논문에서는 지역좌표에 기반을 둔 궤적 데이터 처리 방법을 제안한다. 지역 좌표는 강체변형에 불변한 특징이 있으며, 얼굴모양의 국지적인 변화를 의미하여, 궤적 데이터처리에 효과적으로 활용 될 수 있음을 보였다. 또한 제안한 방법을 활용하여 애니메이션을 제작해 실제 제작 환경에 적용 가능함을 보였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제6권12호
• /
• pp.3315-3337
• /
• 2012

Crowded scenes analysis is a challenging topic in computer vision field. How to detect diverse motion patterns in crowded scenarios from videos is the critical yet hard part of this problem. In this paper, we propose a novel approach to mining motion patterns by utilizing motion information during both long-term period and short interval simultaneously. To capture long-term motions effectively, we introduce Motion History Image (MHI) representation to access to the global perspective about the crowd motion. The combination of MHI and optical flow, which is used to get instant motion information, gives rise to discriminative spatial-temporal motion features. Benefitting from the robustness and efficiency of the novel motion representation, the following motion pattern mining is implemented in a completely unsupervised way. The motion vectors are clustered hierarchically through automatic hierarchical clustering algorithm building on the basis of graphic model. This method overcomes the instability of optical flow in dealing with time continuity in crowded scenes. The results of clustering reveal the situations of motion pattern distribution in current crowded videos. To validate the performance of the proposed approach, we conduct experimental evaluations on some challenging videos including vehicles and pedestrians. The reliable detection results demonstrate the effectiveness of our approach.

• 한국정보통신학회논문지
• /
• 제16권4호
• /
• pp.872-878
• /
• 2012

본 논문은 사람 팔의 움직임 감지 시스템을 제안한다. 기존의 모션측정 장비들은 고가이거나 혹은 특수한 장소가 따로 필요하였다. 또한 측정한 데이터들을 필요한 데이터로 변환하는 추가적인 시간도 필요하였다. 본 논문에서는 사람이 입을 수 있으며 가격이 매우 저렴하고 추가적인 데이터 변환이 필요 없는 시스템을 개발 하였다.

• 한국컴퓨터그래픽스학회논문지
• /
• 제6권3호
• /
• pp.9-15
• /
• 2000

본 논문에서는 모션 캡쳐 시스템을 위한 센서 퓨전 방법을 제안한다. 제안된 시스템에서는 두가지 종류의 센서가 상호보완적으로 사용된다. 연기자의 팔에 부착된 12개의 광학식마커를 보조하기 위해서 4개의 자기식 센서(마커)가 팔 상박과 손등에 부착된다. 광학식 마커는 장애물에 의하여 가려질 수 있기 때문에 광학식 센서의 정보는 불완전할 수 있으며 이 경우 자기식 센서 정보가 불연속인 광학식 센서의 정보를 이어주는데 사용된다. 센서 신호간의 관계를 모델링하기 위해 시스템 동정화 방법을 사용한다. 입 출력 데이터로부터 동적 시스템들을 구성하고 시스템 동정화 방법을 사용하여 후보 모델들로부터 가장 좋은 모델을 결정한다. 제안된 방법은 현재 간단한 신호처리 기법을 사용하고 있는데, 추후 연구에선 Wiener나 Kalman 필터를 사용한 새로운 방법을 제안하고자 한다.

• 한국컴퓨터정보학회논문지
• /
• 제15권4호
• /
• pp.65-74
• /
• 2010

본 논문에서는 기존의 광학식 모션 캡처에서 생길 수 있는 마커들 간의 간섭이나 복잡한 시스템 구성으로 인한 시스템 설치의 복잡성 문제들을 해결하기 위해, 2차원 위치정보를 제공하는 단일 카메라와 특정부위의 방향정보를 제공하는 가속도센서와 자이로 센서로 구성된 동작센서를 융합하여 간편한 모션 캡처를 실현하는 새로운 기법을 제안한다. 본 논문의 동작 인식은 크게 영상기반 위치 정보와 동작센서기반 방향 정보의 융합을 통해 이루어진다. 영상은 보이는 부위에 장착된 컬러마커의 위치를 기준점으로 제공하고, 동작센서들은 각 패지의 이동방향과 속도를 측정하여 영상에서 제공하는 마커들의 3차원 포즈정보를 알아 낼 수 있다. 제안하는 시스템은 사람동작의 측정에 필요한 최소한의 센서정보를 사용함으로써 시스템의 구성과 센서의 설치가 매우 간단하며 경제적이라는 장점을 갖는다. 이러한 장점은 다양한 실험을 통해 검증하였다.

• 한국실내디자인학회논문집
• /
• 제13권3호
• /
• pp.187-196
• /
• 2004

The purpose of this study is to develope'a digital space design process using user's motion data' through a theoretical and experimental study. In the progress of developing a developing of design process, this study was concentrated on searching a digital method applying user's interactive reflections. As introducing a concept of space form being generated by user's experiences, we proposed'a digital design process using user's motion data'. In the experimental stage, user's motion data were extracted and transferred as digital information by user behavior analysis, optical motion capture system, immersive VR system, 3D softwares com computer programming. As the result of this study, another useful digital design process was embodied by building up a digital form-transforming method using 3D softwares providing internal algorithm. This study would be meaningful in terms of attempting a creative and interactive digital space design method, avoiding dehumanization of existing ones through the theoretical study and the experimental approach.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2003년도 학술회의 논문집 정보 및 제어부문 B
• /
• pp.507-510
• /
• 2003

In this paper, optical sensor system using PSD(Position Sensitive Detection) is proposed to obtain the three dimensional position of moving markers attached to human body. To find the coordinates of an moving marrer with stereo vision system, two different sight rays of an moving marker are required. Usually, those are acquired with two optical sensors synchronized at the same time. PSD sensor is used to measure the position of an incidence light in real-time. To get the three-dimension position of light source on moving markers, a conventional camera calibration method are used. In this research, we realized a low cost motion capture system. The proposed system shows high three-dimension measurement accuracy and fast sampling frequency.

• 제어로봇시스템학회논문지
• /
• 제20권1호
• /
• pp.70-77
• /
• 2014

This paper proposes a novel localization algorithm based on ego-motion which used Lucas-Kanade Optical Flow and warping image obtained through fish-eye lenses mounted on the robots. The omnidirectional image sensor is a desirable sensor for real-time view-based recognition of a robot because the all information around the robot can be obtained simultaneously. The preprocessing (distortion correction, image merge, etc.) of the omnidirectional image which obtained by camera using reflect in mirror or by connection of multiple camera images is essential because it is difficult to obtain information from the original image. The core of the proposed algorithm may be summarized as follows: First, we capture instantaneous $360^{\circ}$ panoramic images around a robot through fish-eye lenses which are mounted in the bottom direction. Second, we extract motion vectors using Lucas-Kanade Optical Flow in preprocessed image. Third, we estimate the robot position and angle using ego-motion method which used direction of vector and vanishing point obtained by RANSAC. We confirmed the reliability of localization algorithm using ego-motion based on fisheye warping image through comparison between results (position and angle) of the experiment obtained using the proposed algorithm and results of the experiment measured from Global Vision Localization System.

• 한국CDE학회논문집
• /
• 제17권1호
• /
• pp.62-70
• /
• 2012

This paper proposes a novel discomfort assessment method for truck ingress and egress motions based on the maximum-voluntary-contraction (MVC) ratios of muscles obtained by biomechanical analysis of human musculoskeletal models. In this study, a human motion to enter and exit a truck cabin with different types and heights of footsteps is first measured using an optical motion capture system and load sensors. Next, in a biomechanical analysis system, a human musculoskeletal model with contacting conditions on footsteps and handles is modeled, and then joint torques and muscles forces are calculated by inverse dynamics of the musculoskeletal model with the motion data. Finally, the MVC ratios for the muscles are calculated and their statistical values are used as the measure of discomfort. To ensure the feasibility of our method, subjective discomfort levels have been investigated through the participants' experiments and questionnaires and compared to the results of our method. Comparing to the existing methods based on joint angles or torques, our approach provide a more essential criterion for discomfort because it is based on the muscle contraction by which an active human motion is basically generated.

• 제어로봇시스템학회논문지
• /
• 제20권8호
• /
• pp.868-874
• /
• 2014

This paper proposes a novel mapping algorithm in Omni-directional Vision SLAM based on an obstacle's feature extraction using Lucas-Kanade Optical Flow motion detection and images obtained through fish-eye lenses mounted on robots. Omni-directional image sensors have distortion problems because they use a fish-eye lens or mirror, but it is possible in real time image processing for mobile robots because it measured all information around the robot at one time. In previous Omni-Directional Vision SLAM research, feature points in corrected fisheye images were used but the proposed algorithm corrected only the feature point of the obstacle. We obtained faster processing than previous systems through this process. The core of the proposed algorithm may be summarized as follows: First, we capture instantaneous $360^{\circ}$ panoramic images around a robot through fish-eye lenses which are mounted in the bottom direction. Second, we remove the feature points of the floor surface using a histogram filter, and label the candidates of the obstacle extracted. Third, we estimate the location of obstacles based on motion vectors using LKOF. Finally, it estimates the robot position using an Extended Kalman Filter based on the obstacle position obtained by LKOF and creates a map. We will confirm the reliability of the mapping algorithm using motion estimation based on fisheye images through the comparison between maps obtained using the proposed algorithm and real maps.