• 한국멀티미디어학회:학술대회논문집
• /
• 한국멀티미디어학회 2004년도 춘계학술발표대회논문집
• /
• pp.326-329
• /
• 2004

Content adaptation is an effective solution to support the quality of service over multimedia services over heterogeneous environments. This paper deals with the accuracy and the real-time requirement, two important issues in making decision on content adaptation. From our previous problem formulation, we propose an optimal algorithm and a fast approximation based on the Viterbi algorithm of dynamic programming. Through extensive experiments, we show that the proposed algorithms can enable accurate adaptation decisions, and especially they can support the real-time processing.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권11호
• /
• pp.2720-2736
• /
• 2013

Multi-view face detection has become an active area for research in the last few years. In this paper, a novel multi-view human face detection algorithm based on improved real Adaboost is presented. Real Adaboost algorithm is improved by weighted combination of weak classifiers and the approximately best combination coefficients are obtained. After that, we proved that the function of sample weight adjusting method and weak classifier training method is to guarantee the independence of weak classifiers. A coarse-to-fine hierarchical face detector combining the high efficiency of Haar feature with pose estimation phase based on our real Adaboost algorithm is proposed. This algorithm reduces training time cost greatly compared with classical real Adaboost algorithm. In addition, it speeds up strong classifier converging and reduces the number of weak classifiers. For frontal face detection, the experiments on MIT+CMU frontal face test set result a 96.4% correct rate with 528 false alarms; for multi-view face in real time test set result a 94.7 % correct rate. The experimental results verified the effectiveness of the proposed approach.

• 대한인간공학회지
• /
• 제18권3호
• /
• pp.27-39
• /
• 1999

The purpose of this study was to approve the capability of human sensibility evaluation based on physiological responses and real-time subjective assessments. Three well-trained healthy human subjects were participated in the experiments. We measured physiological responses such as Heart Rate Variability(HRV), Galvanic Skin Response(GSR) and skin temperature under rest and visual stimulation conditions, respectively. Off-time subjective assessments were recorded before and after visual stimulations. Real-time subjective assessments were recorded during visual stimulations. The results of physiological responses and off-time and real-time subjective assessments were quantified and compared. The results showed that the correlation between physiological responses and real-time subjective assessments was high (83%) for both the positive and negative visual stimulation. The correlation between the physiological responses and off-time subjective assessments was high (83%) for the negative visual stimulation but was low (15%) for the positive visual stimulation. Although the current observation is preliminary and requires more careful experimental study, it appears that the correlation between real-time subjective assessment and physiological responses is higher than that of the off-time subjective assessment and physiological responses.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권8호
• /
• pp.3981-4004
• /
• 2019

This paper proposes a novel method for locating objects in real space from a single remote image and measuring actual distances between them by automatic detection and perspective transformation. The dimensions of the real space are known in advance. First, the corner points of the interested region are detected from an image using deep learning. Then, based on the corner points, the region of interest (ROI) is extracted and made proportional to real space by applying warp-perspective transformation. Finally, the objects are detected and mapped to the real-world location. Removing distortion from the image using camera calibration improves the accuracy in most of the cases. The deep learning framework Darknet is used for detection, and necessary modifications are made to integrate perspective transformation, camera calibration, un-distortion, etc. Experiments are performed with two types of cameras, one with barrel and the other with pincushion distortions. The results show that the difference between calculated distances and measured on real space with measurement tapes are very small; approximately 1 cm on an average. Furthermore, automatic corner detection allows the system to be used with any type of camera that has a fixed pose or in motion; using more points significantly enhances the accuracy of real-world mapping even without camera calibration. Perspective transformation also increases the object detection efficiency by making unified sizes of all objects.

• ETRI Journal
• /
• 제37권2호
• /
• pp.251-261
• /
• 2015

We present a novel method for real-time automatic license plate detection in high-resolution videos. Although there have been extensive studies of license plate detection since the 1970s, the suggested approaches resulting from such studies have difficulties in processing high-resolution imagery in real-time. Herein, we propose a novel cascade structure, the fastest classifier available, by rejecting false positives most efficiently. Furthermore, we train the classifier using the core patterns of various types of license plates, improving both the computation load and the accuracy of license plate detection. To show its superiority, our approach is compared with other state-of-the-art approaches. In addition, we collected 20,000 images including license plates from real traffic scenes for comprehensive experiments. The results show that our proposed approach significantly reduces the computational load in comparison to the other state-of-the-art approaches, with comparable performance accuracy.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1996년도 Proceedings of the Korea Automatic Control Conference, 11th (KACC); Pohang, Korea; 24-26 Oct. 1996
• /
• pp.346-349
• /
• 1996

In this paper development of a CAD of control systems is introduced which enables us to do not only analysis of control systems, design of controllers but also real-time implementation of controllers. By utilizing this software, the control engineer is able to repeat the procedure of modification of controllers and experiments without recompile to attain better performance. The software also offers the facility to update the parameters of controllers without stopping real-time control, which helps on-line tuning of controllers. If some parameters of the controller is changed on-line, the control input may change discontinuously. It has serious effect on the control systems. A method for on-line tuning of state feedback controller with state observer is proposed and verified through the experiment with an inverted pendulum.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1997년도 한국자동제어학술회의논문집; 한국전력공사 서울연수원; 17-18 Oct. 1997
• /
• pp.748-751
• /
• 1997

The White Beamline of the Pohang Accelerator Laboratory(PAL) consists of main and second slits, a microprobe system, two ion chambers, a video-microscope, and a Si(Li) detector. These machine components must be controlled remotely through computer system to make user experiments precise and speedy. A real-time computer control system was developed to control and monitor these machine components. A VNIEbus computer with OS-9 real-time operating system was used for low-level data acquisition and control. VME I/O modules were used for step motor control and scaler control. The software has modular structure for maximum performance and easy maintenance. We developed database, I/O driver, and control software. We used PC/Window95 for data logging and operator interface. Visual C++ was used graphical user interface programming. RS232C was used for communication between VME and PC.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.1430-1434
• /
• 2004

Recently, Virtual reality parts is applied in various fields of industry. In this paper we developed basic components for virtual robot control system interfaced with real environment. For this, a real robot with virtual interface module is developed and virtual robot of similar image with real robot is created by putting on 3D graphic texture to the real robot. To build an unknown environment to be linked with virtual environment, we proposed a hough transformation based algorithm. Our proposed algorithm consists of navigation module by using fuzzy engine and map building module. Experiments using a developed robot illustrate the method.

• 제어로봇시스템학회논문지
• /
• 제13권11호
• /
• pp.1082-1091
• /
• 2007

In this paper, we developed a new hybrid mobile robot which can climb stairs and go over thresholds by crawl gait with embedded real-time control software. This robot is also categorized into hybrid robot that has advantages of wheeled mobile robot and legged mobile robot, but adopts gait feature of crocodile named belly crawl. We imitated the belly crawl using four legs of 2 DOF, four omni-directional wheels, and embedded control software which controls legs and wheels. This software is developed using RTAI/Linux, real-time drivers. As a result, the new hybrid mobile robot has crawl gait. Using this feature, the new hybrid mobile robot can climb stairs and go over thresholds just by path planning of each leg with size of stairs and thresholds, and computing the movement distance of robot body center without considering stability. The performance of our new hybrid mobile robot is verified via experiments.

• 한국컴퓨터그래픽스학회논문지
• /
• 제12권1호
• /
• pp.21-25
• /
• 2006

This paper proposes a real-time simulation technique for thin shells undergoing large deformation. Thin shells are almost two-dimensional structures visually well approximated as surfaces, such as leaves, paper sheets, hats, aluminum cans. Unfortunately accurate simulation of these structures requires one of the most complex formulations in continuum mechanics, shell theory [4]. Moreover, there has not yet been any work reported to produce visually convincing animation of them while achieving real-time performance. Motivated by discrete shells [5] and modal warping [3], we formulate dynamics of thin shells using mass-spring models instead of finite element models, and then apply the modal warping technique to cope with large rotational deformation of thin shells. Experiments show that the proposed technique runs in real-time, and that it can simulate large bending and/or twisting deformations with acceptable realism.