• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.9
• /
• pp.88-96
• /
• 1996

Computer vision-based gesture recognition systems consist of image segmentation, object tracking and decision. However, it is difficult to segment an object from image for gesture in computer systems because of vaious illuminations and backgrounds. In this paper, we describe a method to learn features for segmentation, which improves the performance of computer vision-based hand-gesture recognition systems. Systems interact with a user to acquire exact training data and segment information according to a predefined plan. System provides some models to the user, takes pictures of the user's response and then analyzes the pictures with models and a prior knowledge. The system sends messages to the user and operates learning module to extract information with the analyzed result.

• Journal of the Institute of Convergence Signal Processing
• /
• v.11 no.4
• /
• pp.263-269
• /
• 2010

This paper describes real-time detection and tracking of moving objects for unmanned visual surveillance. Using images obtained from the fixed camera it detects moving objects within the image and tracks them with displaying rectangle boxes enclosing the objects. Tracking method is implemented on an embedded system which consists of TI DSK645.5 kit and the FPGA board connected on the DSP kit. The DSP kit processes image processing algorithms for detection and tracking of moving objects. The FPGA board designed for image acquisition and display reads the image line-by-line and sends the image data to DSP processor, and also sends the processed data to VGA monitor by DMA data transfer. Experimental results show that the tracking of moving objects is working satisfactorily. The tracking speed is 30 frames/sec with 320x240 image resolution.

• Journal of Digital Convergence
• /
• v.15 no.2
• /
• pp.173-181
• /
• 2017

In this paper, we have studied tracking as a training stage of considering the position and the scale of a person given its previous position, scale, as well as next and forward image fraction. Unlike other learning methods, CNN is thereby learning combines both time and spatial features from the image for the two consecutive frames. We introduce multiple path ways in CNN to better fuse local and global information. A creative shift-variant CNN architecture is designed so as to alleviate the drift problem when the distracting objects are similar to the target in cluttered environment. Furthermore, we employ CNNs to estimate the scale through the accurate localization of some key points. These techniques are object-independent so that the proposed method can be applied to track other types of object. The capability of the tracker of handling complex situations is demonstrated in many testing sequences. The accuracy of the SVM classifier using the features learnt by the CNN is equivalent to the accuracy of the CNN. This fact confirms the importance of automatically optimized features. However, the computation time for the classification of a person using the convolutional neural network classifier is less than approximately 1/40 of the SVM computation time, regardless of the type of the used features.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.39 no.1
• /
• pp.61-69
• /
• 2002

Active stereo vision system allows us to localize a target object by passing only the features of small disparities without heavy computation for identifying the target. This simple method, however, is not applicable to the situations where a distracting background is included or the target and other objects are located on the zero disparity area simultaneously To alleviate these problems, we combined filtering with foveation which employs high resolution in the center of the visual field and suppresses the periphery which is usually less interesting. We adopted an image pyramid or log-polar transformation for foveated imaging representation. We also extracted the stereo disparity of the target by using projection to keep the stereo disparity small during tracking. Our experiments show that log-polar transformation is superior to either an image pyramid or traditional method in separating a target from the distracting background and fairly enhances the tracking performance.

• Journal of Advanced Navigation Technology
• /
• v.25 no.3
• /
• pp.217-222
• /
• 2021

EOTS(Electro-Optical Tracking System) is utilized in acquiring visual information to assess a guided missile's performance. As the missile travels so fast, it is almost impossible for operator to re-capture the lost target. The RADAR or telemetry data are used to re-capture the lost target however facilities to receive real time data is required, which constrains selection of tracking site. Unlike aforementioned data, pre-calculated nominal trajectory can be used without communication facility. This paper proposes a method to predict lost target's state by employing nominal trajectory. Firstly, observed trajectory and nominal trajectory are compared using DTW and current target's state is predicted. The predicted state is used as observation in Kalman filter's correction phase to predict target's next state. The plausibility of the proposed method is verified by applying on actual missile trajectory.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.16 no.3
• /
• pp.117-122
• /
• 2023

Gaze estimation, which automatically recognizes where a user is currently staring, and object detection based on estimated gaze point, can be a more accurate and efficient way to understand human visual behavior. in this paper, we propose a method to detect the objects within the region of interest around the gaze point. Specifically, after estimating the 3D gaze point, a region of interest based on the estimated gaze point is created to ensure that object detection occurs only within the region of interest. In our experiments, we compared the performance of general object detection, and the proposed object detection based on region of interest, and found that the processing time per frame was 1.4ms and 1.1ms, respectively, indicating that the proposed method was faster in terms of processing speed.

• Journal of the Institute of Convergence Signal Processing
• /
• v.13 no.2
• /
• pp.83-90
• /
• 2012

In this paper, an intelligent visual surveillance system based on a NIOS II embedded platform is implemented. By this time, embedded based visual surveillance systems were restricted for a special purpose because of high dependence upon hardware. In order to improve the restriction, we implement a flexible embedded platform, which is available for various purpose of applications. For high speed processing of software based programming, we improved performance of the system which is integrated the SOPC type of NIOS II embedded processor and image processing algorithms by using software programming and C2H(The Altera NIOS II C-To-Hardware(C2H) Acceleration Compiler) compiler in the core of the hardware platform. Then, we constructed a server system which globally manage some devices by the NIOS II embedded processor platform, and included the control function on networks to increase efficiency for user. We tested and evaluated our system at the designated region for visual surveillance.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.653-659
• /
• 2006

다중 객체 추적이란 컴퓨터 비전의 한 분야로, 주어진 비디오 시퀀스 내에서 관심 있는 객체들을 추적하는 것을 말한다. 다중 객체 추적 시스템은 감시 시스템, 사용자 행동 인식, 스포츠 중계, 비디오 회의와 같은 다양한 응용 분야에 핵심 기반 기술로 쓰이고 있어 그 중요성이 매우 크다. 본 논문은 감시 목적의 다중 객체를 추적하는 방법에 대하여 다룬다. 감시 시스템의 특성상, 객체의 외관이나 움직임 등에 대한 가정을 하기가 어렵다. 따라서 본 논문에서는 크기, 색, 형태 같은 객체의 단순하고 직관적인 외관 특성을 이용하면서도, 객체들끼리 부분적으로 혹은 완전히 겹쳐졌을 때에도 객체들의 위치를 적절히 추적할 수 있는 방법을 제안한다. 본 논문에서 제안하는 방법은 객체들의 경로에 대한 정보를 유지하는데 그래프 구조를 이용한다. 그래프를 확장하고, 제거하여 영상에 대한 정보를 추론한다. 크게 보면 객체들을 영역 레벨, 객체 레벨 두 단계에 걸쳐 추적한다. 영역 레벨에서는 각 객체들이 있을 수 있을만한 영역에 대한 가설을 세우고, 객체 레벨에서는 각 가설에 대한 검증을 한다. 제안된 방법은 직관적인 정보만을 이용하여 서로 다른 형태의 객체를 빠르게 추적할 수 있음을 보여준다. 다만 객체의 외관 정보만을 이용하였기 추적하기 때문에, 객체가 다른 객체에 의해 완전히 가려진 채 또다시 다른 객체와 겹쳐지면, 정확한 추적이 되지 않는다. 이를 해결하기 위해서는 객체가 겹쳐졌을 때, 그 관계에 대한 정보를 모아야 하는데 이는 향후 연구를 통해 해결하고자 한다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2009.01a
• /
• pp.158-163
• /
• 2009

In recent years, there has been increased interest in characterizing and extracting 3D information from 2D images for human tracking and identification. In this paper, we propose a single view-based framework for robust estimation of height and position. In the proposed method, 2D features of target object is back-projected into the 3D scene space where its coordinate system is given by a rectangular marker. Then the position and the height are estimated in the 3D space. In addition, geometric error caused by inaccurate projective mapping is corrected by using geometric constraints provided by the marker. The accuracy and the robustness of our technique are verified on the experimental results of several real video sequences from outdoor environments.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.11
• /
• pp.39-46
• /
• 2012

The development of the computer widens the expressive ways for the nature objects and the scenes. The cutting edge computer graphics technologies effectively create any images we can imagine. Although the computer graphics plays an important role in filming and video production, the status of the domestic contents production industry is not favorable for producing and research all at the same time. In digital composition, the match moving stage, which composites the captured real sequence with computer graphics image, goes through many complicating processes. The camera tracking process is the most important issue in this stage. This comprises the estimation of the 3D trajectory and the optical parameter of the real camera. Because the estimating process is based only on the captured sequence, there are many errors which make the process more difficult. In this paper we propose the method for correcting the tracking data. The proposed method can alleviate the unwanted camera shaking and object bouncing effect in the composited scene.