• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.2
• /
• pp.106-114
• /
• 2017

In this paper, we propose the method of designing an extended kalman filter in order to accurately measure the position of the spatial-phase system using sensor fusion. We use the quaternion as a state variable in expressing the attitude of an object. Then, the attitude of rigid body can be calculated from the accelerometer and magnetometer by applying the Gauss-Newton method. We estimate the changes of state by using the measurements obtained from the gyroscope, the quaternion, and the vision informations by ARVR_SDK. To increase the accuracy of estimation, we designed and implemented the extended kalman filter, which showed excellent ability to adjust and compensate the sensor error. As a result, we could experimentally demonstrate that the reliability of the attitude estimation value can be significantly increased.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.11
• /
• pp.1131-1135
• /
• 2010

This paper presents a method to capture the posture of the human lower-limbs on the 3D space by using a single camera and a planar mirror. The system estimates the pose of the camera facing the mirror by using four coplanar IR markers attached on the planar mirror. After that, the training space is set up based on the relationship between the mirror and the camera. When a patient steps on the weight board, the system obtains relative position between patients' feet. The markers are attached on the sides of both legs, so that some markers are invisible from the camera due to the self-occlusion. The reflections of the markers on the mirror can partially resolve the above problem with a single camera system. The 3D positions of the markers are estimated by using the geometric information of the camera on the training space. Finally the system estimates and visualizes the posture and motion of the both legs based on the 3D marker positions.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.12
• /
• pp.1528-1534
• /
• 2019

Recently, due to the popularization of 3D depth cameras, new researches and opportunities have been made in research conducted on RGB images, but estimation of human hand pose is still classified as one of the difficult topics. In this paper, we propose a robust estimation method of human hand pose from various input 3D depth images using a learning algorithm. The proposed approach first generates a skeleton-based hand model and then aligns the generated hand model with three-dimensional point cloud data. Then, using a random forest-based learning algorithm, the hand pose is strongly estimated from the aligned hand model. Experimental results in this paper show that the proposed hierarchical approach makes robust and fast estimation of human hand posture from input depth images captured in various indoor and outdoor environments.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.1
• /
• pp.155-161
• /
• 2021

Recently, researches for military purposes such as precision tracking and mission completion using UAVs have been actively conducted. In particular, if the posture information of the leading UAV is estimated and the mission UAV uses this information to follow in stealth and complete its mission, the speed of the posture information estimation of the guide UAV must be processed in real time. Until recently, research has been conducted to accurately estimate the posture information of the leading UAV using image processing and Kalman filters, but there has been a problem in processing speed due to the sequential processing of the processing process. Therefore, in this study we propose a way to improve processing speed by applying methods that the image processing area is limited to the ROI area including the object, not the entire area, and the continuous processing is distributed to OpenMP-based multi-threads and processed in parallel with thread synchronization to estimate attitude information. Based on the experimental results, it was confirmed that real-time processing is possible by improving the processing speed by more than 45% compared to the basic processing, and thus the possibility of completing the mission can be increased by improving the tracking and estimating speed of the mission UAV.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.10a
• /
• pp.973-975
• /
• 2016

In this paper, we propose a selective extended Kalman filter based accurate pose estimation of the rigid body using a sensor fusion method. The pose of a rigid body can be estimated roughly by the Gauss-Newton method using the acceleration data and geomagnetic data, which can be refined with vision information and the gyro sensor information. However strong external interference noise makes the rough pose estimation difficult. In this paper, according to the measurement level of the external interference noise, the extended Kalman filter selectively uses mostly vision and gyro sensor information to increase the estimation credibility under strong interference noise environment.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.7
• /
• pp.155-166
• /
• 2012

This paper proposes a new method of estimating the hand pose through the Mean-Shift tracking algorithm using the fusion of color and depth information for marker-less augmented reality. On marker-less augmented reality, the most of previous studies detect the hand region using the skin color from simple experimental background. Because finger features should be detected on the hand, the hand pose that can be measured from cameras is restricted considerably. However, the proposed method can easily detect the hand pose from complex background through the new Mean-Shift tracking method using the fusion of the color and depth information from 3D sensor. The proposed method of estimating the hand pose uses the gravity point and two random points on the hand without largely constraints. The proposed Mean-Shift tracking method has about 50 pixels error less than general tracking method just using color value. The augmented reality experiment of the proposed method shows results of its performance being as good as marker based one on the complex background.

• Journal of Advanced Navigation Technology
• /
• v.26 no.4
• /
• pp.235-243
• /
• 2022

In this paper, in the case of an inertial navigation system, the posture estimation error in the initial alignment due to vertical deflection is analyzed. Posture estimation error due to DOV was theoretically analyzed based on the speed and posture error of INS. Simulations were performed to verify the theoretical grinding, and the results were in good agreement. For example, in the case of η=20", an alignment error of ϕ_N=0.00287°, ϕ_U=0.00196° occurred, and in the case of 𝜉=20", an error of ϕ_E= -0.00286° occurred. Through this, it was confirmed that the vertical posture error caused by the DOV occurred as a coupling characteristic of the INS posture error. It has been shown that an additional posture error may occur due to the DOV, which was not considered in the existing INS alignment, which means that correction for the DOV must be considered when applying high-precision INS.

• Journal of Korean Society of Transportation
• /
• v.21 no.2
• /
• pp.155-164
• /
• 2003

현재 국내에서는 교통사고 재현 방법이 노면 흔적물에 의존하여 이루어지고 있고, 노면 흔적물이 없는 경우 교통사고 재현은 불가능하게 된다. 이러한 단점을 보완하기 위해 외국에서는 충돌모델을 이용한 교통사고 재현프로그램을 활용하고 있는 추세이다. 현재 우리나라에서 가장 많이 사용하는 교통사고 재현 프로그램 PC-CRASH를 활용하기 위해서는 각 사고마다 알맞은 충돌특성 인자 값을 사용자가 직접 입력하여 사용하여야 하나, 이를 활용할 수 있는 활용자료가 부족한 실정이다. 본 연구는 국내에서 실제 발생된 교통사고 사례에 대해 교통사고 재현 프로그램인 PC-CRASH와 충돌에 영향을 미치는 충돌 인자들로 교통사고를 재현하였고, 차량의 최종위치간 거리와 자세에 어떠한 영향을 미치는가를 알아보았다. 또한 실제 역 해석을 하기 위해 차량의 최종위치와 자세만으로 각각의 사례에 적절한 인자값을 추정할 수 있는 회귀식을 구성하였고, 통계적으로 신뢰성을 검증하였다. 사고 재현에 필요한 주요 충돌인자들의 초기치 설정시 추정식을 이용할 경우, 사고 재현 프로그램 활용 시, 시간 단축 효과를 프로그램 내부에 있는 유전자 알고리즘의 반복횟수로 추정식의 통계적 검증을 하였다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2011.10a
• /
• pp.923-926
• /
• 2011

본 연구에서는 움직이는 물체가 있는 비디오에서 검출된 전경 영상(실루엣)을 토대로 사람을 추적하고 추적된 사람의 실루엣 형상을 통하여 활동성을 인식하는 실시간 감시 시스템에 적용 가능한 사람의 행동을 인식하고 분석하고자 한다. 전경에서 블랍(사람)을 검출하는 방법은 기존에 연구했던 차영상을 이용하였고, 검출된 블랍을 대상으로 사람임을 판단하고 사람인 경우 검출된 블랍의 실루엣을 이용한 기존의 자세 추정 기법에 추가적으로 4가지 특징들을 추가하여 사람의 행동을 분석한다. 각 파라미터들은 임계치를 통하여 구분하였다. 본 논문에서는 사람의 행동은 크게 네 가지의 경우로 {Standing, Bending/Crawling, Laying down, Sitting} 분류한다. 제안된 특징 파라미터들을 추가한 방법은 기존의 실루엣 기반의 자세 추정 기법만을 사용하는 것보다 좀더 높은 인식율을 보여주었다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.35-36
• /
• 2022

본 논문은 딥러닝 사람 자세 추정 모델이 사람의 관절 키포인트를 예측하는데 관절의 2차원 면적에 의해 키포인트별 𝜎, 즉, 표준 편차를 가지는 가우시안 커널(Gaussian Kernel)을 예측하는 방법을 제안한다. 각 관절 키포인트에 대해 다른 𝜎를 가지는 정답 히트맵(Ground Truth Heatmap)과 제안한 Gaussian Mixture Block를 모델에 추가해서 관절의 크기를 맞는 히트맵을 예측한다.