• Journal of Korea Multimedia Society
• /
• v.20 no.7
• /
• pp.986-993
• /
• 2017

Object Tracking is a technique for tracking moving objects over time in a video image. Using object tracking technique, many research are conducted such a detecting dangerous situation and recognizing the movement of nearby objects in a smart car. However, it still remains a challenging task such as occlusion, deformation, background clutter, illumination variation, etc. In this paper, we propose a novel deep visual object tracking method that can be operated in robust to many challenging task. For the robust visual object tracking, we proposed a Convolutional Neural Network(CNN) which shares weight of the convolutional layers. Input of the CNN is a three; first frame object image, object image in a previous frame, and current search frame containing the object movement. Also we propose a method to consider the motion of the object when determining the current search area to search for the location of the object. Extensive experimental results on a authorized resource database showed that the proposed method outperformed than the conventional methods.

• Journal of the Institute of Convergence Signal Processing
• /
• v.9 no.1
• /
• pp.18-23
• /
• 2008

In this paper, we propose the input image selection-method to improve the reconstructed high-resolution (HR) image quality. To obtain ideal super-resolution (SR) reconstruction image, all input images are well-registered. However, the registration is not ideal in practice. Due to this reason, the selection of input images with low registration error (RE) is more important than the number of input images in order to obtain good quality of a HR image. The suitability of a candidate input image can be determined by using statistical and restricted registration properties. Therefore, we propose the proper candidate input Low Resolution(LR) image selection-method as a pre-processing for the SR reconstruction in automatic manner. In video sequences, all input images in specified region are allowed to use SR reconstruction as low-resolution input image and/or the reference image. The candidacy of an input LR image is decided by the threshold value and this threshold is calculated by using the maximum motion compensation error (MMCE) of the reference image. If the motion compensation error (MCE) of LR input image is in the range of 0 < MCE < MMCE then this LR input image is selected for SR reconstruction, else then LR input image are neglected. The optimal reference LR (ORLR) image is decided by comparing the number of the selected LR input (SLRI) images with each reference LR input (RLRI) image. Finally, we generate a HR image by using optimal reference LR image and selected LR images and by using the Hardie's interpolation method. This proposed algorithm is expected to improve the quality of SR without any user intervention.

• Journal of Mechanical Science and Technology
• /
• v.18 no.8
• /
• pp.1388-1400
• /
• 2004

Problems with the control, oscillatory motion and compliance of pneumatic systems have prevented their widespread use in advanced robotics. However, their compactness, power/weight ratio, ease of maintenance and inherent safety are factors that could be potentially exploited in sophisticated dexterous manipulator designs. These advantages have led to the development of novel actuators such as the McKibben Muscle, Rubber Actuator and Pneumatic Artificial Muscle Manipulators. However, some limitations still exist, such as a deterioration of the performance of transient response due to the changes in the external inertia load in the pneumatic artificial muscle manipulator. To overcome this problem, a switching algorithm of the control parameter using a learning vector quantization neural network (LVQNN) is newly proposed. This estimates the external inertia load of the pneumatic artificial muscle manipulator. The effectiveness of the proposed control algorithm is demonstrated through experiments with different external inertia loads.

• Journal of the Korean Society for Precision Engineering
• /
• v.30 no.3
• /
• pp.292-299
• /
• 2013

The measurement in the manufacturing process of curved ship hull plates still depends on wooden templates as a standard instrument. The metrology-enabled automation in the shipbuilding process has been challenged instead of line measurement with wooden templates. The developed measurement system consists of a CCD camera, multiple structured laser sources and 3-DOF motion device. The system carries out measurement of curved profiles for large scale plates by an optical triangulation method. The results of experiment conducted in a manufacturing shop demonstrate the accurate and robust performance.

• Journal of Institute of Control, Robotics and Systems
• /
• v.13 no.7
• /
• pp.624-630
• /
• 2007

The dynamic window approach(DWA) is a well known technique for reactive collision avoidance. It shows safe and efficient performance in real-world experiments. However, a robot can get stuck in local minima because no information about the connectivity of the free space is used to determine the motion. The global DWA can solve this problem of local minima by adding a navigation function. Even with the global DWA, it is still difficult for a robot to execute an abrupt change in its direction, for example, entering from the corridor to a doorway. This paper proposes a modified global DWA using the included angles of waypoints extracted from an optimal path. This scheme enables the robot to decelerate in advance before turning into the doorway. Therefore the robot can reach the goal position more safely and efficiently at high speeds.

• Journal of Aerospace System Engineering
• /
• v.10 no.1
• /
• pp.15-20
• /
• 2016

The optimal estimation of a general continuous-discrete system can be achieved through the solution of the Fokker-Planck equation and the Bayesian update. Due the high nonlinearity of the equation of motion of the system and the measurement model, it is necessary to linearize the both equation. To avoid linearization, the filter based on Fokker-Planck equation is designed. with the unscented transformation update mechanism, in which the associated Fokker-Planck equation was solved efficiently and accurately via discrete quadrature and the measurement update was done through the unscented transformation update mechanism. This filter based on the Direct Quadrature Moment of Method(DQMOM) and the unscented transformation update is applied to the bearing only target tracking problem. The proposed filter can still provide more accurate estimation of the state than those of the extended Kalman filter especially when measurements are sparse. Simulation results indicate that the advantages of the proposed filter based on the DQMOM and the unscented transformation update make it a promising alternative to the extended Kalman filter.

• International journal of advanced smart convergence
• /
• v.9 no.2
• /
• pp.179-184
• /
• 2020

Recently, webtoons-based animations have been expanding in Korea, where the role of mobile devices is growing, and media videos such as "YouTube" and "Tik-Tok," which have shorter screening and production cycles than traditional feature films, dramas, and animations, are attracting attention and being enjoyed by public. In order to produce animation content that fits the trend of this shortening video and production cycle, efficiency must be increased not only in story but also in production. Production methods and production technologies need changes to increase efficiency. Animation using motion capture, which is highly production-efficient, is widely used in movies that shows realistic movements, but still has little use in producing cartoon-style animations with exaggerated movements. We analyzed the production method of 2D animation and CG animation and applied the result to CG animation to increase the efficiency of production and production. The methods of production through such experiments are expected to help improve the efficiency of producing animation content that is suitable for the latest trends such as webtoon animation.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.5
• /
• pp.896-901
• /
• 2008

This paper addresses the foot force distribution problem for quadruped robots with a failed leg. The quadruped robot has fault-tolerant straight-line gaits with one leg in locked-joint failure, and has discontinuous motion with respect to the robot body. The proposed method is operated in two folds. When the robot body stands still, we use the feature that there are always three supporting legs, and by incorporating the theory of zero-interaction force, we calculate the foot forces analytically without resort to any optimization technique. When the robot body moves, the conventional pseudo-inverse algorithm is applied to obtain the foot forces for supporting legs. Simulation results show the validity of the proposed scheme.

• Journal of the Korean Institute of Navigation
• /
• v.3 no.1
• /
• pp.1-17
• /
• 1979

It is very important for both naval architects and ship's officers to know the maneuvering characteristics of their ships. As the abilities of a rudder which controlls a ship can be determined clearly by analyzing the results of Kempf's zig-zag maneuver and directional stability of a ship also known by Dieudonn spiral maneuver, the importance of turning test which takes much time is recently apt to be neglected. But because the test can be executed comparatively more simply than any other maneuvering tests, it gives some informations on the directional stability, and turning characteristics may be expressed simply by the results of the test, it is still often performed. In this paper several assumptions are made to simplify the turning motion of a ship. The equations of initial transient phase, the radius ofsteady turning circle, and the center of the steady turning point are derived by using the hydrodynamic derivatives. And then the approximate method of drawing the turning circle geometrically is suggested.

• Proceedings of the KIEE Conference
• /
• 2001.10a
• /
• pp.30-32
• /
• 2001

There are many difficult problems in analyzing the flow characteristics in a high voltage circuit breaker such as shock wave and complex geometries, which may be either static or in relative motion. Although a variety of mesh generation techniques are now available, the generation of meshes around complicated, multi-component geometries like a gas circuit breaker is still a tedious and difficult task for the computational fluid dynamics. This paper presents the CFD program for analyzing the compressible flow fields in a high voltage gas circuit breaker using the Cartesian cut-cell method based on the CFD-CAD integration, which can achieve the accurate representation of the geometry designed by a CAD tools. This technique is frequently satisfied, and it will be almost universally so in the future, as the CFD-CAD traffic increase.