• Korean Journal of Optics and Photonics
• /
• v.19 no.4
• /
• pp.270-275
• /
• 2008

This paper addresses occluded object reconstruction and recognition with computational integral imaging (II). Integral imaging acquires and reconstructs target information in the three-dimensional (3D) space. The reconstruction is performed by averaging the intensities of the corresponding pixels. The distance to the object is estimated by minimizing the sum of the standard deviation of the pixels. We adopt principal component analysis (PCA) to classify occluded objects in the reconstruction space. The Euclidean distance is employed as a metric for decision making. Experimental and simulation results show that occluded targets are successfully classified by the proposed method.

• Journal of the Korea Computer Graphics Society
• /
• v.24 no.3
• /
• pp.33-41
• /
• 2018

We present a secondary camera optimization method that helps the user exploring 3D virtual environment to precisely observe possible collisions between objects. The first role of our secondary camera is to automatically detect the area with the greatest possible collision. The second role is to show the detected area from a new angle of view that the current main camera cannot show. However, as the shapes of target objects are complex, the shape of the empty space between objects is also complex and narrow. It means that the space for the secondary camera position is highly constrained and its optimization can be very difficult. To avoid this difficulty and increase the efficiency of the optimization, we first compute a bisector surface between two target objects. Then, we limit the domain of the secondary camera's position on the bisector surface in the optimization process. To verify the utility of our method, we built a demonstration program in which the user can explore in a 3D virtual world and interact with objects by using a hand motion recognition device and conducted a user study.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2013.11a
• /
• pp.54-57
• /
• 2013

최근 다양한 방송 및 영상 분야에서 사람의 행동을 인식하여는 연구들이 많이 이루어지고 있다. 영상은 다양한 형태를 가질 수 있기 때문에 제약된 환경에서 유용한 템플릿 방법들보다 특징점에 기반한 연구들이 실제 사용자 환경에서 더욱 관심을 받고 있다. 특징점 기반의 연구들은 영상에서 움직임이 발생하는 지점들을 찾아내어 이를 3차원 패치들로 생성한다. 이를 이용하여 영상의 움직임을 히스토그램에 기반한 descriptor(서술자)로 표현하고 학습기반의 판별기(classifier)로 최종적으로 영상 내에 존재하는 행동들을 인식하였다. 그러나 단일 판별기를 이용한 다양한 영상 인식을 수용하기에는 힘들다. 최근에 이를 개선하기 위하여 다중 판별기를 활용한 연구들이 영상 판별 및 물체 검출 영역에서 사용되고 있다. 따라서 본 논문에서는 행동 인식을 위하여 support vector machine과 spare representation을 이용한 decision-level fusion 방법을 제안하고자 한다. 제안된 논문의 방법은 영상에서 특징점 기반의 descriptor를 추출하고 이를 각각의 판별기를 통하여 판별 결과들을 획득한다. 이 후 학습단계에서 획득된 가중치를 활용하여 각 결과들을 융합하여 최종 결과를 도출하였다. 본 논문에 실험에서 제안된 방법은 기존의 융합 방법보다 높은 행동 인식 성능을 보여 주었다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.5
• /
• pp.537-542
• /
• 2011

A mirror neuron is a neuron fires both when an animal acts and when the animal observes the same action performed by another. We propose a method of 3D reconstruction for occluded object motion tracking like Mirror Neuron System to fire in hidden condition. For modeling system that intention recognition through fire effect like Mirror Neuron System, we calculate depth information using stereo image from a stereo camera and reconstruct three dimension data. Movement direction of object is estimated by optical flow with three-dimensional image data created by three dimension reconstruction. For three dimension reconstruction that enables tracing occluded part, first, picture data was get by stereo camera. Result of optical flow is made be robust to noise by the kalman filter estimation algorithm. Image data is saved as history from reconstructed three dimension image through motion tracking of object. When whole or some part of object is disappeared form stereo camera by other objects, it is restored to bring image date form history of saved past image and track motion of object.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.24 no.6
• /
• pp.703-709
• /
• 2015

In this work, a robot aimed at grapping and delivering an object by using a simple finger-pointing command from a hand- or arm-handicapped person is introduced. In this robot system, a Leap Motion sensor is utilized to obtain the finger-motion data of the user. In addition, a Kinect sensor is also used to measure the 3D (Three Dimensional)-position information of the desired object. Once the object is pointed at through the finger pointing of the handicapped user, the exact 3D information of the object is determined using an image processing technique and a coordinate transformation between the Leap Motion and Kinect sensors. It was found that the information obtained is transmitted to the robot controller, and that the robot eventually grabs the target and delivers it to the handicapped person successfully.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.5
• /
• pp.60-72
• /
• 1996

In this paper, a enw model-base three-dimensional (3-D) object recognition mehtod using hopfield network is proposed. To minimize deformation of feature values on 3-D rotation, we select 3-D shape features and 3-D relational features which have rotational invariant characteristics. Then these feature values are normalized to have scale invariant characteristics, also. The input features are matched with model features by optimization process of hopjfield network in the form of two dimensional arrayed neurons. Experimental results on object classification and object matching with the 3-D rotated, scale changed, an dpartial oculued objects show good performance of proposed method.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.1
• /
• pp.55-63
• /
• 1998

This paper presents a new maching algorithm by Hopfield Neural Network for 3-D object recognition. In the proposed method, a model object is represented by a set of polygons in a single coordinate. And each polygon is described by a set of features; feature attributes. In case of 3-D object recognition, the scale and poses of the object are important factors. So we propose a strategy for 3-D object recognition independently to its scale and poses. In this strategy, the respective features of the input or the model objects are changed to the startegical constants when they are compared with one another. Finally, we show that the proposed method has a robustness through the results of experiments which included the classification of the input objects and the matching sequence to its 3-D rotation and scale.

• Proceedings of the KIEE Conference
• /
• 1990.11a
• /
• pp.470-473
• /
• 1990

In this thesis, the recognition method of the movement of an 3D object is presented. The information about the movement of a 3D object is used to recognize the object. There are 2 kinds of movements which are translation and rotation. A difference picture is obtained from a sequence of images of a moving object or a scene which is taken by a monocular stationary observer. The 3D movement of an object is recognized by the Artificial Neural Network(ANN) using the difference picture.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.34C no.9
• /
• pp.84-101
• /
• 1997

Inferring and recognizing 3D objects form a 2D occuluded image has been an important research area of computer vision. The octree model, a hierarchical volume description of 3D objects, may be utilized to generate projected images from arbitrary viewing directions, thereby providing an efficient means of the data base for 3D object recognition. We present a fast algorithm of finding the 4 pairs of feature points to estimate the viewing direction. The method is based on matching the object contour to the reference occuluded shapes of 49 viewing directions. The initially best matched viewing direction is calibrated by searching for the 4 pairs of feature points between the input image and the image projected along the estimated viewing direction. Then the input shape is recognized by matching to the projectd shape. The computational complexity of the proposed method is shown to be O(n$^{2}$) in the worst case, and that of the simple combinatorial method is O(m$^{4}$.n$^{4}$) where m and n denote the number of feature points of the 3D model object and the 2D object respectively.

• Journal of Korea Multimedia Society
• /
• v.15 no.8
• /
• pp.963-976
• /
• 2012

'Gesture' except for voice is the most intuitive means of communication. Thus, many researches on how to control computer using gesture are in progress. User detection and tracking in these studies is one of the most important processes. Conventional 2D object detection and tracking methods are sensitive to changes in the environment or lights, and a mix of 2D and 3D information methods has the disadvantage of a lot of computational complexity. In addition, using conventional 3D information methods can not segment similar depth object. In this paper, we propose object detection and tracking method using Depth Projection Map that is the cumulative value of the depth and motion information. Simulation results show that our method is robust to changes in lighting or environment, and has faster operation speed, and can work well for detection and tracking of similar depth objects.