• 대한전자공학회논문지SP
• /
• 제44권6호
• /
• pp.75-83
• /
• 2007

본 논문은 영상 기반의 사람의 자세 추정에 대하여 다룬다. 특히 사람이 걷는 동안 카메라는 사람의 측면을 관찰하고 있다고 가정한다. 사람의 자세 추정의 문제는 인간-컴퓨터 상호 작용이나 지능형 감시 시스템을 위해 연구가 되는 분야이며, 본 논문에서는 일반적인 보행 상황에서 감시 시스템 또는 위치 추적, 자세 인식에 응용할 수 있는 알고리즘을 제시한다. 이 분야의 최근의 연구동향은 마코프 네트워크를 이용하여 신체 부분들의 위치나 움직임의 관계를 조건부 독립으로 가정하여 다루고 있다. 이러한 방법들의 경우 신체를 십여 개의 부분들로 모델링하고, 연결된 신체들의 관계를 고려하여 자세를 추정한다. 본 논문에서는 이러한 방법을 응용하여 모델을 단순화하고, 더 나아가 손쉽게 사람의 자세를 파악할 수 있는 방법을 제시한다. 이를 위해 신체 부분들이 독립적임을 가정하여 그 위치를 찾은 후에, 모션 캡쳐 데이터로부터 얻은 신체 부분들의 움직임 간의 관계를 고려하여 자세를 수정하여 주었다. 사람의 신체를 찾기 위해 edge matching을 이용하였으며, 그 과정에서 신체 부분의 edge 성분의 방향성을 강조하기 위해 Anisotropic Gaussian Filter를 사용하였다. 신체의 부분이 가려지는 경우, 모델의 silhouette을 이용하여 가려지는 부분에 대해 추가의 matching cost를 부여함으로써 occlusion 시에도 신체의 부분을 찾을 수 있도록 하였다.

• 전기학회논문지
• /
• 제56권9호
• /
• pp.1655-1664
• /
• 2007

This paper presents the problem of controlling asynchronous sequential machines in the presence of input disturbances, which may be also regarded as an adversary in a game theoretic setting. The main objective is to develope a new methodology for including unpredictable behavior of input disturbance into models of asynchronous machines. The input disturbance, representing uncontrollable noise input, is embedded into a new model of asynchronous machines in form of input/state finite state machines. It is shown that the proposed modeling preserves the fundamental model and well-pose of asynchronous machines. The reachability matrix, an important performance index of asynchronous machines, is also adapted according to input disturbance and will be used for constructing corrective controllers in the companion paper.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2005년도 추계학술대회 학술발표 논문집 제15권 제2호
• /
• pp.359-362
• /
• 2005

본 논문에서는 홍채 인식을 할 때 여러 가지 자세 변화에 민감한 홍채 패턴을 일정한 기준에 따라 항상 고정된 형태로 추출하기 위해 눈꺼풀의 윤곽을 검출하여 눈의 모양을 바로잡는 방법을 소개한다. 이와 더불어 효과적인 홍채 영역 검출을 위한 정확한 동공의 경계 측정과 공막 경계 측정을 위한 새로운 방법을 제시한다. Template Matching과 Mean Shape을 이용하여 여러 가지 다양한 눈의 형태와 눈썹의 영향 때문에 판단이 까다로운 눈꺼풀의 경계를 검출하였다. 동공 경계의 자세한 검출은 Hough Transform을 이용하였고 공막의 경계는 최소 자승법을 이용하였다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2001년도 추계학술대회 학술발표 논문집
• /
• pp.299-302
• /
• 2001

The statistical analysis of the feature extraction and the neural networks are proposed to recognize a human face. In the preprocessing step, the normalized skin color map with Gaussian functions is employed to extract the region of face candidate. The feature information in the region of the face candidate is used to detect the face region. In the recognition step, as a tested, the 120 images of 10 persons are trained by the backpropagation algorithm. The images of each person are obtained from the various direction, pose, and facial expression. Input variables of the neural networks are the geometrical feature information and the feature information that comes from the eigenface spaces. The simulation results of$.$10 persons show that the proposed method yields high recognition rates.

• 로봇학회논문지
• /
• 제9권3호
• /
• pp.170-177
• /
• 2014

This paper proposes a novel upward-looking camera-based global localization using a ceiling image map. The ceiling images obtained through the SLAM process are integrated into the ceiling image map using a particle filter. Global localization is performed by matching the ceiling image map with the current ceiling image using SURF keypoint correspondences. The robot pose is then estimated by the coordinate transformation from the ceiling image map to the global coordinate system. A series of experiments show that the proposed method is robust in real environments.

• Journal of information and communication convergence engineering
• /
• 제9권2호
• /
• pp.207-211
• /
• 2011

This paper presents a method for facial feature representation and recognition from the Covariance Matrices of the Gabor-filtered images. Gabor filters are a very powerful tool for processing images that respond to different local orientations and wave numbers around points of interest, especially on the local features on the face. This is a very unique attribute needed to extract special features around the facial components like eyebrows, eyes, mouth and nose. The Covariance matrices computed on Gabor filtered faces are adopted as the feature representation for face recognition. Geodesic distance measure is used as a matching measure and is preferred for its global consistency over other methods. Geodesic measure takes into consideration the position of the data points in addition to the geometric structure of given face images. The proposed method is invariant and robust under rotation, pose, or boundary distortion. Tests run on random images and also on publicly available JAFFE and FRAV3D face recognition databases provide impressively high percentage of recognition.

• 한국기계가공학회지
• /
• 제3권4호
• /
• pp.7-12
• /
• 2004

Inspection and analysis are essential process to determine whether a completed product is in given specification or not. Analysis of products with very complicated shape is difficult to carry out direct comparison between inspected coordinate and designed coordinates. So process called as matching or registrations is needed to solve this problem. By defining error between two coordinates and minimizing the error, registration is done. Registration consists of translation, rotation and scale transformations. Error must be defined to express feature of inspected product. In this paper, registration algorithm is developed to determine pose of sub-frame at assembly with body of automobile by defining error between two coordinates considering geometric feature of sub-frame.

• 로봇학회논문지
• /
• 제4권3호
• /
• pp.169-176
• /
• 2009

Low-cost sensors have been widely used for mobile robot navigation in recent years. However, navigation performance based on low-cost sensors is not good enough to be practically used. Among many navigation techniques, building of an accurate map is a fundamental task for service robots, and mapping with low-cost IR sensors was investigated in this research. The robot's orientation uncertainty was considered for mapping by modifying the Bayesian update formula. Then, the data association scheme was investigated to improve the quality of a built map when the robot's pose uncertainty was large. Six low-cost IR sensors mounted on the robot could not give rich data enough to align the range data by the scan matching method, so a new sample-based method was proposed for data association. The real experiments indicated that the mapping method proposed in this research was able to generate a useful map for navigation.

• 로봇학회논문지
• /
• 제9권2호
• /
• pp.96-103
• /
• 2014

Global localization is one of the essential issues for mobile robot navigation. In this study, an indoor global localization method is proposed which uses a Kinect sensor and a monocular upward-looking camera. The proposed method generates an environment map which consists of a grid map, a ceiling feature map from the upward-looking camera, and a spatial feature map obtained from the Kinect sensor. The method selects robot pose candidates using the spatial feature map and updates sample poses by particle filter based on the grid map. Localization success is determined by calculating the matching error from the ceiling feature map. In various experiments, the proposed method achieved a position accuracy of 0.12m and a position update speed of 10.4s, which is robust enough for real-world applications.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2006년도 하계종합학술대회
• /
• pp.423-424
• /
• 2006

This paper describes a 3D face recognition using different devices for 3D faces and input faces which include several different pose. Before the recognition stage, through the EC-SVD, all data have to be preprocessed and normalized. At recognition stage, we propose the multi-point signature method for measuring facial surface information. And we use the root mean square error for matching. From the experiment results, we have 92.5% recognition rate.