Loading [MathJax]/jax/output/CommonHTML/jax.js
  • Title/Summary/Keyword: 문제해결 지도

Search Result 8,322, Processing Time 0.034 seconds

The Risk Detection Technique for Visually Impaired Pedestrians using YOLOv8 on GloSea6 (GloSea6에서 YOLOv8을 활용한 시각장애인 보행자 위험 감지 기법)

  • Nam-Gyu Kim;Jin-Seong Park;Sung-Yeon Kim;Hye-Seong Park;Sung-Wook Chung
    • The Journal of Korea Institute of Information, Electronics, and Communication Technology
    • /
    • v.17 no.6
    • /
    • pp.451-461
    • /
    • 2024
  • Weather changes have a significant impact on the safety of personal mobility devices, with conditions such as rain and snow being identified as major causes of increased accident rates. To address these issues, the necessity of safety systems utilizing high-performance weather prediction models like GloSea has been emphasized. The rapid proliferation of personal mobility devices worldwide has greatly influenced urban traffic environments and pedestrian safety. In South Korea, the growing use of these devices has heightened concerns about the safety of pedestrians, particularly those with visual impairments. This study proposes a system that uses the YOLOv8 model for object detection, depth estimation, and Kalman filter-based tracking techniques to quantitatively evaluate the trajectories and risk levels of objects. Experimental results show that the custom YOLOv8 model achieved an average mAP50 of 0.774, with a high performance of mAP50 0.717 for the personal mobility device class. High-risk objects are visualized with red boundaries, and when combined with GloSea-based weather data, the system can contribute to predicting and warning of accident risks under varying weather conditions. In the future, this system is expected to support safe navigation for visually impaired individuals through integration with voice feedback.

Research to improve the Efficiency through Maturity Model System of Manufacturing Plant (제조공장의 성숙도 모델 시스템을 통한 효율성 향상 연구)

  • Jinseok Kim;Younggon Kim
    • The Journal of the Institute of Internet, Broadcasting and Communication
    • /
    • v.24 no.6
    • /
    • pp.99-105
    • /
    • 2024
  • Manufacturing plants around the world are facing rapid changes in the wake of the Fourth Industrial Revolution. South Korea is also rapidly changing with government support. However, while building an intelligent factory, there is a lack of evaluation factors and quantitative research on manufacturing plants, insufficient leveling of selected checklists, and insufficient planning and on-site verification of improvement work. In order to solve these problems, this paper presented a maturity model using work weighting using a practice model in the business process of a manufacturing plant. The basic manufacturing plant's business process was divided into 6 business management categories, and the maturity model was classified into Step A, Step B, and Step C levels, and each level was given a maturity level of 1 to 5. Through this, each task function was selected according to the detailed level of maturity, and each task was weighted for analysis and evaluation by level, and improvements to business functions and processes were proposed. In addition, the system was built through UML-based design to implement the Maturity Model System. It is expected that this will enable quantitative evaluation, level diagnosis, and improvement items using work weights. Furthermore, it is expected that it will be possible to establish medium- to long-term improvement plans and derive short-term improvement work through level diagnosis centered on field strategies.

Mandibular full arch rehabilitation of a patient with severely worn dentition after completion of maxillary full arch rehabilitation (상악 편악 수복을 선행한 심한 마모를 가진 환자에서의 하악 편악 수복 증례)

  • Young-Min Kim;Seung-Geun Ahn;Jung-Jin Lee;Yeon-Hee Park;Kyoung-A Kim;Jae-Min Seo
    • Journal of Dental Rehabilitation and Applied Science
    • /
    • v.40 no.4
    • /
    • pp.279-290
    • /
    • 2024
  • Wear of teeth occurs due to various reasons, and full mouth rehabilitation with an accurate diagnosis of the patient can restore the occlusal relationship, aesthetics, and facial appearance and prevent the progression of wear. Restoration of worn dentition must be performed throughout the entire upper and lower arch. If restoration is performed only on the dentition of single arch, various problems may occur. The patient in this case had generalized wear of the maxillary and mandibular dentition and required full-mouth rehabilitation. However, due to the patient's request, only the maxillary arch was rehabilitated first, and later, due to complications, the mandibular arch was rehabilitated. Purpose of this case report is to consider the problems when only the maxillary full arch was rehabilitated in a patient with generalized tooth wear, and how these were resolved during the mandibular full arch rehabilitation process.

Development and Evaluation of Shielding with Mixed Iron- oxide-copper Filament using 3D printing (3D 프린팅 기술을 이용한 산화철-구리 혼합 필라멘트 차폐체 개발 및 차폐율 평가)

  • Nareoyng Shin;Seong-gwan Nam;JiSu Kang;GeonJu Lee;HuiMin Jang;Myeong-Seong Yoon;Dong-Kyoon Han
    • Journal of the Korean Society of Radiology
    • /
    • v.18 no.6
    • /
    • pp.651-662
    • /
    • 2024
  • This study developed a lead-free shielding material using a filament made from a mixture of iron oxide and copper to address the toxicity issues associated with lead-based shielding. After creating the filaments, various thicknesses of shielding materials were printed using a FDM 3D printer. Shielding performance and dose measurements were taken by varying the tube voltage and current with a diagnostic X-ray generator, and data analysis was performed using SPSS (p < 0.05). The results showed that the iron oxide-copper mixed filament shielding exhibited better performance than a 0.25 mmPb lead-equivalent protection tool at thicknesses of 8 mm or more, except for the conditions of 120 kVp and 20 mAs. This research demonstrates that the mixed filament shielding can mitigate the drawbacks of lead while providing comparable shielding effectiveness, suggesting its potential as foundational data for further studies on lead-free shielding materials.

A Study on Secure Node Memory Allocation in ROS Composition (ROS Composition에서의 안전한 노드 메모리 할당에 관한 연구)

  • Jiwon Seo
    • The Transactions of the Korea Information Processing Society
    • /
    • v.13 no.12
    • /
    • pp.669-675
    • /
    • 2024
  • The Robot Operating System (ROS) has become a critical platform for developing advanced robotic systems across various fields, including smart homes, autonomous driving, and industrial automation. By modularizing complex robotic systems, ROS enables nodes to cooperate effectively while adopting a structure where each node operates as an independent process, ensuring fault isolation. However, the use of inter-process communication (IPC) introduces performance degradation and resource inefficiencies, particularly in scenarios involving frequent message exchanges or large-scale data processing. To address these challenges, ROS2 introduced the Composition feature, allowing multiple nodes to execute within a single process. While this approach improves communication efficiency by leveraging shared memory, it also raises potential security risks due to the shared address space. Specifically, loose memory access controls between nodes increase the likelihood of security vulnerabilities, which existing mechanisms such as Secure ROS (SROS) struggle to mitigate effectively. This paper proposes a novel method to enhance security and improve memory management in ROS Composition environments. By independently managing the memory segments of each node, the proposed mechanism prevents data interference and ensures secure memory access. Experimental results demonstrate that the method maintains stable performance with a low overhead of 3% to 10%, effectively enhancing the security of ROS Composition environments.

A Study on an Automatic Model Creation Tool for Applying Structured UML Models in Software Development (소프트웨어 개발에 구조화된 UML 모델을 적용하기 위한 자동 모델 생성 도구에 관한 연구)

  • Seungmo Jung;Woojin Lee
    • The Transactions of the Korea Information Processing Society
    • /
    • v.13 no.12
    • /
    • pp.683-690
    • /
    • 2024
  • Recently, large-scale software development has been using the highly readable Unified Modeling Language (UML) development method. The use of standardized UML models in software development improves software quality by resolving unclear communication. However, in existing software development, a code-centric development method is applied rather than a model-centric development method. As a result, problems such as increased model creation work time arise because UML models must be manually handled when developing model-centric software using existing software. In addition, depending on the developer's understanding of the model, the ability to use modeling tools, and the complexity of the software structure, the time required to create the model increases further. The increase in model creation work time is a factor that increases the overall software development time. Therefore, this paper proposes an automatic model creation tool for structurally applying UML models to the development of naval combat system software. The Automatic Model Creation Tool provides features that automatically generate model structures and UML models needed for modeling tasks. Using the method proposed in this paper, it has the advantage of structurally applying UML models through automation functions and efficiency in reducing model creation work time.

Efficient Scheduling for Inference of Complex Neural Network Models on Multi-GPU Systems (다중 GPU 시스템에서의 복잡한 신경망 모델 추론을 위한 효율적인 스케줄링)

  • Sunwook Jung;Seongju Lee;Beom Woo Kang;Yongjun Park
    • The Transactions of the Korea Information Processing Society
    • /
    • v.13 no.11
    • /
    • pp.604-618
    • /
    • 2024
  • The main challenge facing recent complex neural network models, which have shown competitive accuracy, is their efficient deployment in multi-GPU systems. The complex inter-layer dependences of the neural network models combined with the variable data communication overhead of multi-GPU systems make it almost impossible to achieve a fair performance gain under manual scheduling. To address this problem, we propose a new layer-scheduling approach called NN Maestro, which generates an efficient parallel execution strategy for multi-GPU systems that minimizes the data communication overhead, thereby improving the inference latency for complex neural network models. NN Maestro evaluates the advantages of multi-GPU scheduling using a pre-trained SVM classifier and calculates the scheduling order of layers based on Topological Sort and Significance Cost. Then, NN Maestro selects the optimal GPU by comparing Placement Costs and generates the final scheduling result by grouping layers for parallel execution. On various multi-GPU configurations (2 2080Ti, 4 V100, and 4 A6000 GPUs), NN Maestro achieves up to 1.67x of performance improvement over the baseline.

A Comparative Analysis of General Characteristics, Stress Level, Lifestyle, and Scalp Hair Management Awareness of Hair Losses and Non-Talted Hair Losses (탈모인과 비탈모인의 일반적 특성과 스트레스 정도, 생활 습관, 두피모발 관리인식에 대한 비교분석)

  • Jeon Yunhee
    • Journal of Advanced Technology Convergence
    • /
    • v.3 no.4
    • /
    • pp.99-105
    • /
    • 2024
  • The purpose of this study was to analyze how stress, hair management perception, and management experience are related to hair loss, and to suggest hair loss prevention and scalp management measures. As a result of the study, people with hair loss tended to have a higher stress index and a mainly oily scalp compared to non-metallic hair. They often felt that their hair was thinner and their hair volume decreased from 3 to 5 years ago. Respondents' hair care habits such as shampoo, dry, and dyeing were investigated, and dry or frequent dyeing with hot wind was the main cause of hair damage. Lack of confidence was pointed out as the biggest problem as the social impact of hair loss. As the main place for hair loss management, Vitalmoin tended to prefer scalp specialized care stores and hair loss people tended to prefer hospitals. Based on this, it emphasized the importance of stress, rest and sleep, and lifestyle to prevent hair loss. As scalp management measures, it was recommended to use shampoo suitable for the oily scalp, practice proper cleaning and drying methods, and regular care and professional counseling. This study is meaningful in spreading the right awareness of hair loss and scalp care and in suggesting practical solutions.

Reliability of a Questionnaire for Evaluation of Dry Mouth Symptoms (구강건조증 증상 평가를 위한 설문지의 신뢰도에 관한 연구)

  • Lee, Jeong-Yun;Lee, Young-Ok;Kho, Hong-Seop
    • Journal of Oral Medicine and Pain
    • /
    • v.30 no.4
    • /
    • pp.383-389
    • /
    • 2005
  • Xerostomia is defined as a subjective complaint of dry mouth that may be perceived when there is insufficient mucosal wetting. However, the diagnosis and treatment of xerostomia is not that simple because of the fact that the subjective awareness of dry mouth is not always correlated with a diminution in the flow of saliva and there is always a difference between individuals in salivary flow rates needed for normal oral function. In the present study, the aim was to develop a questionnaire to evaluate the dry mouth symptoms and to analyze its reliability and usability as a diagnostic and analytic tool for xerostomia. The questionnaire which consists of 6 Visual Analogue Scale(VAS) type questions to evaluate the subjective oral dryness and 4 questions to evaluate behavior to avoid oral dryness was developed and administered twice with 1 week's interval to the healthy 88 young adults without dry mouth symptoms(44 males and 44 females; mean age was 25.6±3.1 years in male and 24.3±2.1 years in female). The results were as follows. 1 The intraclass correlation coefficients of 6 questions to evaluate the subjective oral dryness were as significantly high as 0.767 for the degree of oral dryness at other times of the day, 0.850 for the amount of saliva in the mouth, and 0.791 for the degree of effect on daily life due to oral dryness and as high as 0.563 for the degree of oral dryness at night or on awakening, 0.674 for the degree of oral dryness during eating, and 0.641 for the degree of difficulty in swallowing foods. 2. Cronbach's alpha value of 6 questions was 0.982. It can be concluded that the series of questions to evaluate the subjective oral dryness has high internal consistency. 3. Cohen's kappa values of 4 questions to evaluate behavior to relieve oral dryness were as significantly high as 0.850 for the frequency of keeping a glass of water at the bedside and as high as 0.506 the frequency of awakening during sleeping due to oral dryness, 0.419 for the frequency of sipping liquids to aid in eating dry foods, and 0.407 for the frequency of using a candy or chewing gum due to oral dryness. From the results, it can be concluded that the questionnaire consisting of 6 VAS type questions to evaluate the subjective oral dryness and 4 questions to evaluate behavior to relieve oral dryness has reliability of good to excellent level, and that the series of 6 VAS type questions has significantly high internal consistency to evaluate the subjective oral dryness.