• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.25 no.3
• /
• pp.347-355
• /
• 2001

A heat transfer analysis in a ventilation chamber of Plasma Display Panel(PDP) has been conducted. The process requirement is to precisely follow prescribed temperature trajectory while maintaining temperature uniformity for each panel. Firstly, experiment in a test chamber has been carried out and the results are compared with the unsteady 3D numerical data. Reasonable agreement was found, which suggested that the employed numerical model had its credibility in actual PDP ventilation processes. On this ground, a tact-type heating/cooling system was analyzed. The panel temperature in the 40$^{\circ}C$ tact-type system was more uniform than that in the 80$^{\circ}C$ one. For improving the uniformity of panel temperature, relocation of ventilation head to the rear part and inlet flow control are required. Comparison of full simulation of a cart and simplified simulation of one panel indicates the optimized panel pitch can also be predicted.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2004.11a
• /
• pp.594-597
• /
• 2004

본 연구에서는 진공 인라인 실장 기술을 이용하여 제작한 플라즈마 디스플레이 패널(PDP)의 전기적 광학적 특성을 측정하여, 일반적인 실장 방법을 이용한 PDP의 특성과 비교 분석하였다. 본 실험에 사용된 패널은 Screen Printer를 이용한 상 하부전극과 하판 유전체, 상판 투명유전체, 격벽 및 E-Beam Evaporation 방법을 이용하여 증착한 MgO 보호막으로 이루어져 있으며, 분위기 온도 $430^{\circ}C$, Ne-Xe(4%) 400[torr]압력 하에서 실장하였다. 높은 분위기 온도로 인하여 MgO에 Crack이 발생하였으나 지속적인 연구를 진행하여 최적의 실장 조건을 확립할 수 있었다. 이러한 진공 인라인 실장 기술은 추가적인 Annealing 공정이 필요하지 않아 공정의 단축을 모색할 수 있으며, MgO의 수화를 제거함으로써 일반적인 실장 방법을 이용한 패널보다 더 우수한 전기적 광학적 특성을 얻을 수 있었다.

• International journal of advanced smart convergence
• /
• v.10 no.4
• /
• pp.110-116
• /
• 2021

This paper proposed a real-time earlobe detection system using deep learning on the web. Existing deep learning-based detection methods often find independent objects such as cars, mugs, cats, and people. We proposed a way to receive an image through the camera of the user device in a web environment and detect the earlobe on the server. First, we took a picture of the user's face with the user's device camera on the web so that the user's ears were visible. After that, we sent the photographed user's face to the server to find the earlobe. Based on the detected results, we printed an earring model on the user's earlobe on the web. We trained an existing YOLO v5 model using a dataset of about 200 that created a bounding box on the earlobe. We estimated the position of the earlobe through a trained deep learning model. Through this process, we proposed a real-time earlobe detection system on the web. The proposed method showed the performance of detecting earlobes in real-time and loading 3D models from the web in real-time.

• International journal of advanced smart convergence
• /
• v.8 no.1
• /
• pp.65-74
• /
• 2019

3D file formats typically include OBJ (Wavefront file format), STL (STereoLithography), and FBX (Filmbox). Each format has limitations depending on its configuration and usage, and supported formats are different depending on the software application. glTF helps uniform integration of 3D file formats and allows for more efficient transmission of large 3D geometry files by organizing them in a binary format. This paper presents explanation on OBJ, FBX, and STL which are major examples of existing 3D file formats. It also explains the concept and characteristics of glTF and compares its performance with other 3D file formats on the web. The loading time and packets of each 3D file format are measured according to the web browser environment by means of Google Chrome, Firefox and Microsoft Edge. Experimental results show that glTF is the most efficient and that it exhibits the best performance. As to STL, relatively excessive traffic was observed. This study is expected to contribute to reducing rendering time on the web as 3D file formats are used.

• International Journal of Advanced Culture Technology
• /
• v.9 no.2
• /
• pp.86-90
• /
• 2021

Sentiment refers to a person's thoughts, opinions, and feelings toward an object. Sentiment analysis is a process of collecting opinions on a specific target and classifying them according to their emotions, and applies to opinion mining that analyzes product reviews and reviews on the web. Companies and users can grasp the opinions of public opinion and come up with a way to do so. Recently, natural language processing models using the Transformer structure have appeared, and Google's BERT is a representative example. Afterwards, various models came out by remodeling the BERT. Among them, the Facebook AI team unveiled the XLM-R (XLM-RoBERTa), an upgraded XLM model. XLM-R solved the data limitation and the curse of multilinguality by training XLM with 2TB or more refined CC (CommonCrawl), not Wikipedia data. This model showed that the multilingual model has similar performance to the single language model when it is trained by adjusting the size of the model and the data required for training. Therefore, in this paper, we study the improvement of Korean sentiment analysis performed using a pre-trained XLM-R model that solved curse of multilinguality and improved performance.

• International Journal of Internet, Broadcasting and Communication
• /
• v.14 no.3
• /
• pp.22-29
• /
• 2022

Recently, with the development of mobile computing power, mobile-based VR and AR services are being developed. Due to network performance and computing power constraints, VR and AR services using large-capacity 3D content have limitations. A study on an efficient 3D content load method for a mobile device is required. The conventional method downloads all 3D content used for AR services at the same time. In this paper, we propose an active 3D content load according to the user's track. The proposed method is a partitioned 3D object load. Edge servers were installed for each area and connected through the MESH network. Partitioned load the required 3D object in the area referring to the user's location. The location is identified through the edge server information of the connected AP. The performance of the proposed method and the conventional method was compared. As a result of the comparison, the proposed method showed a stable Mobile AR Service. The results of this study, it is expected to contribute to the activation of edge server-based AR mobile services.

• International Journal of Internet, Broadcasting and Communication
• /
• v.14 no.3
• /
• pp.276-284
• /
• 2022

In this paper, we implemented cloud rendering using WebRTC for high-quality AR and VR services. Cloud rendering is an applied technology of cloud computing. It efficiently handles the rendering of large volumes of 3D content. The conventional VR and AR service is a method of downloading 3D content. The download time is delayed as the 3D content capacity increases. Cloud rendering is a streaming method according to the user's point of view. Therefore, stable service is possible regardless of the 3D content capacity. In this paper, we implemented cloud rendering using WebRTC and analyzed its performance. We compared latency of 100MB, 300MB, and 500MB 3D AR content in 100Mbps and 300Mbps internet environments. As a result of the analysis, cloud rendering showed stable latency regardless of data volume. On the other hand, the conventional method showed an increase in latency as the data volume increased. The results of this paper quantitatively evaluate the stability of cloud rendering. This is expected to contribute to high-quality VR and AR services

• International Journal of Internet, Broadcasting and Communication
• /
• v.15 no.4
• /
• pp.215-222
• /
• 2023

Augmented reality (AR), which integrates virtual media into reality, is increasingly utilized across various industrial sectors, thanks to advancements in 3D graphics and mobile device technologies. The IT industry is thus carrying out active R&D activities about AR platforms. Google plays a significant role in the AR landscape, with a focus on ARCore services. An essential aspect of ARCore is the use of anchors, which serve as reference points that help maintain the position and orientation of virtual objects within the physical environment. However, if the accuracy of anchor positioning is suboptimal when running AR content, it can significantly diminish the user's immersive experience. We are to assess the performance of these anchors in this study. To conduct the performance evaluation, virtual 3D objects, matching the shape and size of real-world objects, we strategically positioned ourselves to overlap with their physical counterparts. Images of both real and virtual objects were captured from five distinct camera trajectories, and ARCore's performance was analyzed by examining the difference between these captured images.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.37 no.3
• /
• pp.314-321
• /
• 2024

Low road lighting is a lighting device that complements the shortcomings of existing pillar-type street lights. It is a lighting device that emits light from the side of the road surface and adjusts the luminance of the road surface like a light carpet. In this paper, to achieve full commercialization, we analyzed the luminance of the installed road surface and studied whether lighting could replace existing road lighting. In this study, the LMK (Luminance Measurement Camera) LABSOFT program was used to measure and analyze the surface luminance of road lighting, and the RELUX program was used to evaluate and analyze the simulation performance to determine light-based lighting conditions. A study was conducted to determine whether replacing pillar-type road lighting with low-level road lighting in a real environment would ensure comfortable and safe night vision for drivers at night.

• Korean Journal of Artificial Intelligence
• /
• v.12 no.3
• /
• pp.1-7
• /
• 2024

In this paper, we propose an innovative super-resolution technique to address the issue of reduced accuracy in license plate recognition caused by low-resolution images. Conventional vehicle license plate recognition systems have relied on images obtained from fixed surveillance cameras for traffic detection to perform vehicle detection, tracking, and license plate recognition. However, during this process, image quality degradation occurred due to the physical distance between the camera and the vehicle, vehicle movement, and external environmental factors such as weather and lighting conditions. In particular, the acquisition of low-resolution images due to camera performance limitations has been a major cause of significantly reduced accuracy in license plate recognition. To solve this problem, we propose a Single Image Super-Resolution (SISR) model with a parallel structure that combines Multi-Scale and Attention Mechanism. This model is capable of effectively extracting features at various scales and focusing on important areas. Specifically, it generates feature maps of various sizes through a multi-branch structure and emphasizes the key features of license plates using an Attention Mechanism. Experimental results show that the proposed model demonstrates significantly improved recognition accuracy compared to existing vehicle license plate super-resolution methods using Bicubic Interpolation.