초록 ▼

1. 2단계 수행 중 그래픽스 분야 최고 학술지 “ACM Transactions on Graphics”에 6건의 논문을 게재하였음.
2. 실제 공연에서 사용할 수 있는 동작을 만들어 내는 것에 대한 연구인 “Shadow Theatre: Discovering Human Motion from a Sequence of Silhouettes”을 그래픽스 분야 최대 학회인 “SIGGRAPH 2016”에서 발표하였으며, 각종 보도 자료에서 대중에게 해당 연구결과를 소개하였음.
3. “Push-Recovery Stability of

1. 2단계 수행 중 그래픽스 분야 최고 학술지 “ACM Transactions on Graphics”에 6건의 논문을 게재하였음.
2. 실제 공연에서 사용할 수 있는 동작을 만들어 내는 것에 대한 연구인 “Shadow Theatre: Discovering Human Motion from a Sequence of Silhouettes”을 그래픽스 분야 최대 학회인 “SIGGRAPH 2016”에서 발표하였으며, 각종 보도 자료에서 대중에게 해당 연구결과를 소개하였음.
3. “Push-Recovery Stability of Biped Locomotion”을 그래픽스 분야 최고 학회인 “SIGGRAPH ASIA 2015”에서 발표하였음.
4. 이 외 4편의 논문을 “SIGGRAPH ASIA 2014” 및 “SIGGRAPH 2014”에서 발표하였음.

( 출처 : 보고서 요약서 3p )

Abstract ▼

Purpose
The major goal of this research project is to develop a new way of understanding, analyzing and synthesizing the biological motion of humans and animals. We intend to build a computer simulation model of biological brains and sensory-motor systems in order to understand how biological mot

Purpose
The major goal of this research project is to develop a new way of understanding, analyzing and synthesizing the biological motion of humans and animals. We intend to build a computer simulation model of biological brains and sensory-motor systems in order to understand how biological motions are generated and modulated. We will conduct experiments with biped, quadrupled, and flying creatures to build a unified model of biological motions that can be applied to subjects regardless of their musculoskeletal structures and inhabitation. We will apply our model of biological motion and its simulation method to biped/quadrupled robot control and clinical gait analysis.

contents
From an academy point of view, we will study the biological motion of humans and vertebrates to understand its generation process and reconstruct biological motion in physically based simulation. Our subjects would include humans (biped), iguanas(quadrupled), parrots and pigeons (flying) to capture their biological motion using marker-based optical motion capture, markerless high-speed video, and ground force plates. The motion data thus obtained will be compared to simulation results to verify our model and evaluate the accuracy of the simulation. The practical application at the first phase of our research project includes the control and simulation of biped/quadrupled robots to make them move like a real human/creature while keeping them dynamically balanced. The application at the second phase of the project includes clinical gait analysis based on our model of biological motion and physically based simulation. We intend to develop a system that predicts and visualizes gait patterns subject to surgical operation, impaired(balance, pressure, or vision) sensors, and the use of prostheses.

Developement results
We published two SCI papers in this year. Shadow Theatre: Discovering Human Motion from a Sequence of Silhouettes, a research about making postures that can be used in actual performance, was published in ACM Transactions on Graphics, the top journals in the field of computer graphics. We presented the result in SIGGRAPH 2016, the most prestigious conference in the graphics field.

Furthermore, we presented our research results to the public. A study on making postures that can be used in actual performance was introduced in major media.

Expected Contribution
The robust simulation and control of biological (in particular, biped) motion has been a long-standing problem in robotics and biomechanics. Most existing humanoid robots, such as Honda Asimo, exhibit stereotyped quasi-static walking that looks robotic and unnatural. Achieving dynamic walking and a variety of natural human behaviors in humanoid robots still necessitate significant academic challenges. In practice, biological motion modeling and simulation would allow humanoid and biologically-inspired robots to better imitate a real human/creature.

Through clinical gait analysis based on accurate model and physically based simulation, we expect to better predict and visualize post-operative gait patterns and understand how impaired sensors affect human motion.

( 출처 : SUMMARY 5p )

목차 Contents

• 표지 ... 1
• 제출문 ... 2
• 보고서 요약서 ... 3
• 국문 요약문 ... 4
• SUMMARY ... 5
• 목차 ... 6
• 1. 연구개발 목표 및 내용(연구계획서상에 작성된 목표 및 내용을 기술) ... 7
• 2. 연구 추진전략 및 방법 ... 9
• 2.1. 휴머노이드 로봇 분석 및 제어 ... 12
• 2.2. 정적 인체 운동 분석/시뮬레이션 ... 16
• 2.3. 근골격 모델링 ... 21
• 2.4. 휴머노이드 로봇 제어 ... 27
• 2.5. 동적 인체 운동 분석 ... 31
• 2.6. 동적 인체 운동 시뮬레이션 ... 35
• 2.7. 군집 행위 분석 ... 42
• 2.8. 군집 행위 가시화 ... 50
• 2.9. 실시간 동작 생성 및 실제 공연에 적용 ... 55
• 2.10. 의료 행위 분석 ... 60
• 2.11. 의료 행위 가시화 ... 63
• 3. 주요 연구개발결과 ... 70
• 4. 연구수행에 따른 문제점 및 개선방향 ... 76
• 5. 연구개발성과 현황 ... 76
• 6. 국가과학기술종합정보시스템에 등록한 연구시설․장비 현황 ... 84
• 7. 기타사항 ... 84
• 끝페이지 ... 85