• Proceedings of the KSR Conference
• /
• 2008.11b
• /
• pp.518-523
• /
• 2008

Using network communication, KTX OBCS is KTX's core device which inspects operation status of various electric equipments and supports driving command, train control, train's maintenance, train crew's driving. Since this device is manufactured as specialized system, being dependent on TGV manufacture specification and France's high speed train operation method, actual technic transfer is difficult. Therefore, there are many difficulties, since most of interface signals are connected with OBCS when function improvement, maintenance, various electric equipments(propulsion control device, auxiliary power supply device, door control device etc.) are replaced with localized ones, which are required in operational view after introduction of KTX. Ergo, this study would like to contribute to function improvement and maintenance's efficiency, furthermore to realize technical independence, escaping from dependency of developed foreign technology in high speed train and electric equipments, through KTX OBCS' function analysis and mutual interface information analysis with peripheral device(driver console, PID system, fault display device, various electric equipments).

• Proceedings of the Korean Society Of Semiconductor Equipment Technology
• /
• 2005.09a
• /
• pp.193-196
• /
• 2005

Low resistance Ni thin films for using NiSi formation and metallization by atomic layer deposition (ALD) method have been studied. ALD temperature window is formed between $200^{\circ}C\;and\;250^{\circ}C$ with deposition rate of $1.25{\AA}$/cycle. The minimum resistance of deposited Ni films shows $4.333\;{\Omega}/\square$ on the $SiO_2/Si$ substrate by $H_2$ direct purging process. The reason of showing the low resistance is believed to be due to format ion of the $Ni_3C$ phase by residual carbon in Bis-Ni The deposited film exhibits excellent step coverage in the trench having 1(100 nm) : 16 (1.6 um) aspect ratio.

• Journal of Convergence for Information Technology
• /
• v.11 no.9
• /
• pp.28-34
• /
• 2021

In the case of object detection using deep learning, both accuracy and real-time are required. However, it is difficult to use a deep learning model that processes a large amount of data in a limited resource environment. To solve this problem, this paper proposes an object detection model for small embedded devices. Unlike the general detection model, the model size was minimized by using a structure in which the pre-trained feature extractor was removed. The structure of the model was designed by repeatedly stacking lightweight convolution blocks. In addition, the number of region proposals is greatly reduced to reduce detection overhead. The proposed model was trained and evaluated using the public dataset PASCAL VOC. For quantitative evaluation of the model, detection performance was measured with average precision used in the detection field. And the detection speed was measured in a Raspberry Pi similar to an actual embedded device. Through the experiment, we achieved improved accuracy and faster reasoning speed compared to the existing detection method.

• Journal of Convergence for Information Technology
• /
• v.11 no.10
• /
• pp.45-52
• /
• 2021

In this paper, we proposed an Atrous Residual U-Net (AR-UNet) to improve the segmentation accuracy of semantic segmentation method based on U-Net. The U-Net is mainly used in fields such as medical image analysis, autonomous vehicles, and remote sensing images. The conventional U-Net lacks extracted features due to the small number of convolution layers in the encoder part. The extracted features are essential for classifying object categories, and if they are insufficient, it causes a problem of lowering the segmentation accuracy. Therefore, to improve this problem, we proposed the AR-UNet using residual learning and ASPP in the encoder. Residual learning improves feature extraction ability and is effective in preventing feature loss and vanishing gradient problems caused by continuous convolutions. In addition, ASPP enables additional feature extraction without reducing the resolution of the feature map. Experiments verified the effectiveness of the AR-UNet with Cityscapes dataset. The experimental results showed that the AR-UNet showed improved segmentation results compared to the conventional U-Net. In this way, AR-UNet can contribute to the advancement of many applications where accuracy is important.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.2
• /
• pp.379-384
• /
• 2021

With people's continuous research and exploration in the field of artificial intelligence, more relatively mature artificial intelligence technology is applied in people's daily life. Mirrors are the most commonly used daily necessities in life, and it should be applied to artificial intelligence. The research results of this paper show that the smart mirror designed based on the raspberry pi displays weather, temperature, greetings, and has a human-mirror interaction function. The research method of this paper uses the Raspberry pi 3B + as the core controller and Google Assistant as the intelligent control. When connected to the network via Raspberry Pi's own WiFi, the mirror can automatically display and update time, weather and news information features. You can wake up the Google Assistant using keywords, then control the mirror to play music, remind the time, It implements the function of smart mirror voice interaction. Also, all the hardware used in this study is modular assembly. Later, it is convenient for user to assemble by himself later. It is suitable for market promotion at an affordable price.

• Journal of Convergence for Information Technology
• /
• v.9 no.3
• /
• pp.75-81
• /
• 2019

In this work, we investigated a low temperature polycrystalline silicon (LTPS) thin film transistors fabrication process on polymer layers. Dehydrogenation and activation processes were performed by a furnace annealing at a temperature of $430^{\circ}C$ for 2 hr. The crystallization of amorphous silicon films was formed by excimer laser annealing (ELA) method. The p-type device performance, fabricated by polycrystalline silicon (poly-Si) films, shows a very good performance with field effect mobility of $77cm^2/V{\cdot}s$ and on/off ratio current ratio > $10^7$. We believe that the poly-Si formed by a LTPS process may be well suited for fabrication of poly-Si TFTs for bendable panel displays such as AMOLED that require circuit integration.

• Journal of Convergence for Information Technology
• /
• v.9 no.1
• /
• pp.92-98
• /
• 2019

Due to the development of hologram display technology, various studies are being conducted to provide realistic contents for augmented reality. In the case of the HMD for hologram, since augmented reality objects must be rendered by a small processor, it is necessary to use a low-capacity content. To solve this problem, there is a need for a technique of rendering resources by providing resources through a network. In the case of the existing augmented reality system, there is no problem of contents modulation because the resources are loaded and rendered in the internal storage space. However, when providing resources through the network, security problems such as content tampering and malicious code insertion should be considered. Therefore, in this paper, we propose a network rendering technique applying security techniques to provide augmented reality contents in a holographic HMD device.

• Journal of Conservation Science
• /
• v.37 no.4
• /
• pp.362-369
• /
• 2021

In museums, exhibition content focuses mostly on cultural heritage's historical values and functions, but doing so tends to limit visitors' interest and immersion. To counter this limitation, the study developed an experiential media art exhibition fusing bronze mirrors' traditional production technology and modern conservation science. First, for the exhibition system, scientific cultural heritage contents were projected on the three-dimensional (3D) printed bronze mirror through interactions between motion recognition digital information display (DID) and the projector. Then, a scenario of 17 missions in four stages (production process, corrosion mechanism, scientific analysis and diagnosis, and conservation treatment and restoration) was prepared according to the temporal spectrum. Additionally, various media art effects and interaction technologies were developed, so visitors could understand and become immersed in bronze mirrors' scientific content. A user test was evaluated through the living lab, reflecting generally high levels of satisfaction (90.2 points). Qualitative evaluation was generally positive, with comments such as "easy to understand and useful as the esoteric science exhibition was combined with media art" (16.7%), "wonderful and interesting" (11.7%), and "firsthand experience was good" (9.2%). By combining an esoteric science exhibition centered on principles and theories with visual media art and by developing an immersive directing method to provide high-level exhibition technology, the exhibition induced visitors' active participation. This exhibition's content can become an important platform for expanding universal museum exhibitions on archaeology, history, and art into conservation science.

• Journal of Industrial Convergence
• /
• v.19 no.6
• /
• pp.49-57
• /
• 2021

Digital Signage and multi-view image system are used as the 4th media to deliver stories and information due to their strong immersion. A content image displayed on large Digital Signage is produced with the use of computer graphics, rather than reality image. That is because the images shot for content making have an extremely limited range of production and their limitation to high resolution, and thereby have difficulty being displayed in a large and wide Digital Signage screen. In case of Screen X and Escape that employ the left and right walls of in the center a movie theater as a screen, images are shot with three cameras for Digital Cinema, and are screened in a cinema with multi-view image system after stitching work is applied. Such realistic images help viewers experience real-life content. This research will be able to display high-resolution images on Digital Signage without quality degradation by using the multi-view image making technique of Screen X and Chroma key technique are showed the high-resolution Digital Signage content making method.

• Investigative Magnetic Resonance Imaging
• /
• v.26 no.2
• /
• pp.117-124
• /
• 2022

Purpose: In this study, we provide a way to assess even a slight effect of exercise on trunk-muscle activity. Materials and Methods: Seven healthy male participants (mean age, 24.7 ± 3.2 years; height, 171.2 ± 9.8 cm; and weight, 63.8 ± 11.9 kg) performed 15 sets of an exercise with 20 repetitions of 90° hip and right-knee flexion while lying supine. The exercise intensity was measured using the 10-point Rating of Perceived Exertion Scale after the first and 15th sets of exercises. Although cross-sectional areas and functional T2 mapping using ultrafast imaging (fast-acquired muscle functional magnetic resonance imaging, fast-mfMRI) have been proposed for imaging to evaluate exercise-induced muscle activity in real time, no previous studies have reported on the evaluation of trunk-muscle activity using functional T2 mapping. As a method for assessing trunk-muscle activity, we compared functional T2 mapping using ultrafast imaging (fast-mfMRI) with cross-sectional areas. Results: Although the muscle cross-sectional areas were increased by the exercise, there was no significant difference at rest. On the other hand, for all sets, the changes in T2 were significant compared with those at rest (P < 0.01). These results demonstrate that T2, calculated from fast-mfMRI images can be used to detect even a small amount of muscle activity induced by acute exercise, which was impossible to do with cross-sectional areas. Conclusion: Fast-mfMRI, which can also display functional information with detailed forms, enabled non-invasive real-time imaging for identifying and evaluating the degree of deep trunk-muscle activity induced by exercise.