• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2003.10a
• /
• pp.213-218
• /
• 2003

The equipment of industrial robot in manufacturing and assembly lines has rapidly increased. In order to achieve high productivity and flexibility, it becomes very important to develop the visual feedback control system with Off-Line Programming System(OLPS). We can save much efforts and time in adjusting robots to newly defined workcells by using OLPS. A proposed visual calibration scheme is based on position-based visual feedback. The calibration program firstly generates predicted images of objects in an assumed end-effector position. The process to generate predicted images consists of projection to screen-coordinates, visible range test and construction of simple silhouette figures. Then camera images acquired are compared with predicted ones for updating position and orientation data. Computation of error is very simple because the scheme is based on perspective projection which can be also expanded to experimental results. Computation time can be extremely reduced because the proposed method does not require the precise calculation of tree-dimensional object data and image Jacobian.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.11
• /
• pp.1482-1484
• /
• 2019

We present a method for transmitting 3D object information in real time in a telepresence system. Three-dimensional object information consists of a large amount of point cloud data, which requires high performance computing power and ultra-wideband network transmission environment to process and transmit such a large amount of data in real time. In this paper, multiple users can transmit object motion and facial expression information in real time even in small network bands by using GANs (Generative Adversarial Networks), a non-supervised learning machine learning algorithm, for real-time transmission of 3D point cloud data. In particular, we propose the creation of an object similar to the original using only the feature information of 3D objects using conditional GANs.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.19 no.1
• /
• pp.39-46
• /
• 2024

In this paper, we propose an efficient method for estimating the number of people and their locations using multiple IR-UWB radar sensors. Using three IR-UWB radar sensors in the indoor space, the measured signal from the target is processed to remove the clutter using rejection methods. Then, to further remove the clutter and to determine the presence of the human, the time-frequency image representing the micro-Doppler is obtained and classified by a convolutional neural network. Finally, the system finds the number of human objects and estimates each position in a two-dimensional space. In experiments using the measured data, the system successfully estimated the location and number of individuals with a high accuracy ≈ 88.68 %.

• Journal of the Korean Wood Science and Technology
• /
• v.3 no.1
• /
• pp.3-15
• /
• 1975

How to stabilize wood against shrinking and swelling in variable atmospheric moisture conditions is important to the wood-using industry and a challenge to research. Polyethylene glycol stabilize wood by bulking the fiber. PEG also serve as a chemical seasoning agent, suppress decay in high concentrations, and have slight effect on physical properties, gluing or finishing. The study designed to determine the effect of PEG-400 on the dimensional stabilization of local hardwoods for wood carvings that could supply a greatly expanding tourist trade and making curved furniture parts, lamp stands and other decorative objects, and possible gunstock. The species examined were 6 species, Seo-Namoo (Carpinus laxiflora), Cheungcheung-Namoo (Cornus controversa), Gorosae-Namoo (Acer mono), Karae-Namoo (Juglans mandshurica), Jolcham-Namoo (Quercusserrata) and Sanbud-Namoo (Prunus sargentii), used as block of 5cm thick radially to the grain, 7cm wide tangentially, and 70cm long parallel to the wood grain. All these test piecies were conditioned above the fiber saturation point before impregnation. The stabilization effects were determined for PEG-400 treated woods in a 50 percent solution for 20 days. The following conclusions were obtained. PEG retentions increased with treating time. It was more effective to treat at 60$^{\circ}C$ than at room temperature. In degree of PEG-400 impregnation on species, Cheungcheung-Namoo havinglow specific gravity had the highest retentions, 68.77% but the lowest, 56.33% was shown in Jolcham-Namoo with high specific gravity. Specific gravity of treated wood increased considerably with effectiveness of polymer loading. The increases in specific gravity were 5.36 to 13.16 percent. The highest was Jolcham-Namoo, the lowest Karae-Namoo. On the dimensional stability, a 40 percent of effectiveness of polymer loading was just as effective as 60 percent in reduction in water absorptivity (RWA), antishrinkage efficiency (ASE) and antiswelling efficiency (AE), and from over 60 percent they increased more rapidly. Also species response varied considerably. ASE was 30.12 to 69.97 percent tangentially and 27.86 to 56.37 percent radially, AE 34.06 to 73.76 percent tangentially and 30.11 to 70.12 percent radially, and RWA 42.31 to 65.32 percent. No differences in volume swelling among the 6 species were observed. Its values were ranged from 14.98 to 19.55 percent and also increased with PEG retentions. On the mechanical properties, the strengths very much decreased with PEG-400 loadings as shown in Figure 12; that were 11.41 to 22.90 percent in compression, 21.61 to 34.35 percent in bending and 22.83 to 36.83 percent in tensile strength. PEG retention in cell wall was less than 1 percent and the most of PEG were immersed in cell lumen. Except for Korae-Namoo, effectivenesses of polymer loading were as much high as 61.58 to 75.02 percent. This is believed to be due to the effect of PEG-400 on excellant dimensional stability of treated woods.

• Journal of KIISE
• /
• v.42 no.2
• /
• pp.242-254
• /
• 2015

In this paper, we address the problem of improving the performance of multi-step k-NN search using multi-dimensional indexes. Due to information loss by lower-dimensional transformations, existing multi-step k-NN search solutions produce a large tolerance (i.e., a large search range), and thus, incur a large number of candidates, which are retrieved by a range query. Those many candidates lead to overwhelming I/O and CPU overheads in the postprocessing step. To overcome this problem, we propose two efficient solutions that improve the search performance by reducing the tolerance of a range query, and accordingly, reducing the number of candidates. First, we propose a tolerance reduction-based (approximate) solution that forcibly decreases the tolerance, which is determined by a k-NN query on the index, by the average ratio of high- and low-dimensional distances. Second, we propose a coefficient control-based (exact) solution that uses c k instead of k in a k-NN query to obtain a tigher tolerance and performs a range query using this tigher tolerance. Experimental results show that the proposed solutions significantly reduce the number of candidates, and accordingly, improve the search performance in comparison with the existing multi-step k-NN solution.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.06a
• /
• pp.121-124
• /
• 2007

This paper describes the development of a brickwork game with three-dimensional computer graphics as one of web-based game contents. Client user using web can access and run lava applet program with the independence of hardware system. It consists of graphic user interface module, initial space generation module, event handler module, player control module, and thread control module. It uses 3-D array data structure for the 3-D graphic objects that are located in three-dimensional space for high-speed object searching and sorting. It enhances to compare with predetermined construction in three-dimensional space. We can use the developed racing game to inform game users of information for an advertisement like tourism information, and can apply the proposed 3-D drawing technology to 3-D game graphic engine core.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.2
• /
• pp.486-492
• /
• 2015

Radiation source imaging system is essential for protecting of radiation leakage accidents and minimizing damages from the radioactive materials, and is expected to play an important role in the nuclear plant decommissioning area. In this study, the stereoscopic camera principle was applied to develop a new radiation imaging device technology that can extract the radiation three-dimensional position information. This radiation three-dimensional imaging device (K3-RIS) was designed as a compact structure consisting of a radiation sensor, a CCD camera, and a pan-tilt only. It features the acquisition of stereoscopic radiation images by position change control, high-resolution detection by continuous scan mode control, and stereoscopic image signal processing. The performance analysis test of K3-RIS was conducted for a gamma-ray source(Cs-137) in radiation calibration facility. The test result showed that a performance error with less than 3% regardless of distances of the objects.

• International Journal of Precision Engineering and Manufacturing
• /
• v.9 no.4
• /
• pp.16-21
• /
• 2008

The demand for three-dimensional (3D) shape measurements is increasing in a variety of fields, including the manufacture of molds and dies. The most popular technology for 3D shape measurement is the coordinate measuring machine (CMM) with a contact trigger probe. Although a CMM provides a high degree of accuracy, it is inefficient due to its long measuring time. It also has difficulty measuring soft objects that can be deformed by the touch of the contact probe. In addition, a CMM cannot digitize areas that are difficult to reach, and cannot capture very minute details on the surface of complex parts. For these reasons, optical non-contact measurement techniques are receiving more attention since they eliminate most of the problems associated with contact methods. Laser scanning is emerging as one of the more promising non-contact measurement techniques. This paper describes various acquisition considerations for laser scanning, including the accuracy of the 3D scan data, which depends on the charge-coupled device (CCD) gain and noise. The CCD gain and noise of a 3D laser scanner are varied while keeping the other conditions constant, and the measurement results are compared to the dimensions of a standard model. The experimental results show that a considerable time savings and an optimum degree of accuracy are possible by selecting the proper CCD gain and noise.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.32 no.10
• /
• pp.784-790
• /
• 2008

The feasibility and the accuracy of the correlation coefficient(CC) method for the determination of particle positions along the optical axis in digital particle holography were verified by alidation experiments. A traverse system with capable of high precision was used to move the particle objects by exact known distances between several different positions. The particle positions along the optical axis were calculated by the CC method and compared with their exact values to obtain the errors of the focal plane determination. The tested particles were 2D dots in a calibration target along with different sized glass beads and droplets that reflected and caused a three-dimensional effect. The results show that the CC method can work well for both the 2D dots and the 3D particles. The effect of other particles on the focal plane determination was also investigated. The CC method can locate the focal plane of particles with a high precision, regardless of the existence of other particles.

• Journal of Electrical Engineering and Technology
• /
• v.13 no.5
• /
• pp.2099-2106
• /
• 2018

This study presents a novel approach of discriminative feature vectors based on manifold learning using nonlinear dimension reduction (DR) technique to improve loss function, and combine with the Adversarial examples to regularize the object function for image classification. The traditional convolutional neural networks (CNN) with many new regularization approach has been successfully used for image classification tasks, and it achieved good results, hence it costs a lot of Calculated spacing and timing. Significantly, distrinct from traditional CNN, we discriminate the feature vectors for objects without empirically-tuned parameter, these Discriminative features intend to remain the lower-dimensional relationship corresponding high-dimension manifold after projecting the image feature vectors from high-dimension to lower-dimension, and we optimize the constrains of the preserving local features based on manifold, which narrow the mapped feature information from the same class and push different class away. Using Adversarial examples, improved loss function with additional regularization term intends to boost the Robustness and generalization of neural network. experimental results indicate that the approach based on discriminative feature of manifold learning is not only valid, but also more efficient in image classification tasks. Furthermore, the proposed approach achieves competitive classification performances for three benchmark datasets : MNIST, CIFAR-10, SVHN.