• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.44 no.3
• /
• pp.98-107
• /
• 2007

For high-dimensional pattern recognition, such as face classification, the small number of training samples leads to the Small Sample Size problem when the number of pattern samples is smaller than the number of dimensionality. Recently, various LDA-extensions have been developed, including LDA, PCA+LDA, and Direct-LDA, to address the problem. This paper proposes a method of improving the classification efficiency by increasing the number of (sub)-classes through pre-clustering a training set prior to the execution of Direct-LDA. In LDA (or Direct-LDA), since the number of classes of the training set puts a limit to the dimensionality to be reduced, it is increased to the number of sub-classes that is obtained through clustering so that the classification performance of LDA-extensions can be improved. In other words, the eigen space of the training set consists of the range space and the null space, and the dimensionality of the range space increases as the number of classes increases. Therefore, when constructing the transformation matrix, through minimizing the null space, the loss of discriminatve information resulted from this space can be minimized. Experimental results for the artificial data of X-OR samples as well as the bench mark face databases of AT&T and Yale demonstrate that the classification efficiency of the proposed method could be improved.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.2
• /
• pp.109-115
• /
• 2020

As the maritime targets that threaten allies become lower, smaller, and faster, the need for analysis and modeling of clutter according to sea state increases. Clutter according to the sea state has a great influence on radar performance, such as lowering the probability of detection of low-altitude small maritime targets. In this paper, to analyze the detection performance of a multi function radar for a ship, a sea clutter model suitable for the radar operating environment is selected from several sea clutter models, and analysis of low-altitude, small target detection under a clutter is performed. By using the actual data of the already mounted radar for maritime target detection, four known clutter models have been implemented for each sea state and compared with the actual data. Through this, by selecting a clutter model that best reflects the actual radar environment, reliability of the clutter model is improved. Subsequently, the selected model is used to detect the detectable distance to the low-altitude small target.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.28 no.4
• /
• pp.37-47
• /
• 2024

This study proposes a track maintenance quantity estimation model that considers evolution of rail fatigue damage and ballast settlement based on actual maintenance data from the Gyeongbu high-speed railway, and revises the existing life cycle cost (LCC) model for railway track. Using this model, maintenance quantities and life cycle costs based on different track quality levels are evaluated and discussed. According to the results, it is confirmed that applying the track maintenance quantity estimation model that accounts for rail fatigue damage and ballast settlement allows us to reasonably estimate maintenance costs close to the actual data. The track quality coefficient significantly influences both rail and ballast maintenance quantities, with ballast maintenance having a greater impact than rail maintenance. Additionally, as train speed increases, both rail and ballast maintenance quantities rise. Moreover, a higher track quality coefficient leads to a steeper increase in maintenance quantities with increasing train speed. Consequently, LCC also exhibits a faster growth rate over time with higher track quality coefficients and faster train speeds, resulting from an increased proportion of maintenance costs.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.38 no.4
• /
• pp.375-382
• /
• 2020

As the 4th Industrial Revolution era in worldwide, interest in autonomous vehicles is increasing. but due to recent safety issues such as pedestrian accidents and car accidents, as a technical model for this, the demand for 3D HD maps (High Definition maps) is increasing in including lanes, road markings, road information, traffic lights and traffic signs etc. However, since some complementary points have been continuously raised according to demand, It is necessary to collect the opinions of institutions and companies utilizing HD maps and to improve HD maps. This study was conducted by utilizing the results of the contest for usability verification of HD Maps hosted by the National Geographic Information Institute and organized by the Spatial Information Industry Promotion Institute. For this study, we researched HD maps' layers and codes for HD maps object usability to improve HD maps, constructed HD maps object usability items accordingly, and contested usability verification of HD maps according to the items The contestants conducted verification and analyzed the results. As a result, the most frequently used code for each layer was the flat intersection, and the code showing the highest usage rate was a safety sign. In addition, the use rate of the sub-section and height obstacles was 16.67% and 8.88%, respectively, showing a low ratio. In order to utilize HD maps in the future, this study is expected to require research to continuously collect opinions from customers and improve data objects and data models that are actually needed by customers.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.25 no.3
• /
• pp.165-189
• /
• 2014

This study examined the current practices of Research Data Management (RDM) services recently built and implemented at research university libraries in the U.S. by analyzing the components of the services and the contents presented in their web sites. The study then analyzed the content of web pages describing the services provided by 31 Research Universities/Very High research activity determined based on the Carnegie Classification. The analysis was based on 9 components of the services suggested by previous studies, including (1) DMP support; (2) File organization; (3) Data description; (4) Data storage; (5) Data sharing and access; (6) Data preservation; (7) Data citation; (8) Data management training; (9) Intellectual property of data. As a result, the vast majority of the universities offered the service of DMP support. More than half of the universities provided the services for describing and preserving data, as well as data management training. Specifically, RDM services focused on offering the guidance to disciplinary metadata and repositories of relevance, or training via individual consulting services. More research and discussion is necessary to better understand an intra- or inter-institutional collaboration for implementing the services and knowledge and competency of librarians in charge of the services.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.12 no.6
• /
• pp.44-53
• /
• 2013

Ongoing SWIM(System Wide Information Management) with the United States and European countries as the center is a part of the ASBU(Aviation System Block Upgrade) program improved performance of aeronautical data system in the International Civil Aviation Organization and a core technology of Integrated Aeronautical Data Management Network to elevate service through digitally aeronautical information management. Therefore, in this paper, we analyze SWIM architecture and network applied the concept of SOA(Service Oriented Architecture), and propose methods of implementation transforming applications operating established legacy aeronautical data system into integrated aeronautical data management network through adapter technology. This will allow development of middleware and application suitable for the next generation infrastructure network environment for efficient ATM(Air Traffic Management)and provide timely required information for users.

• Journal of Internet of Things and Convergence
• /
• v.4 no.2
• /
• pp.1-6
• /
• 2018

Techniques for generating new sample data from higher dimensional data such as images have been utilized variously for speech synthesis, image conversion and image restoration. This paper adopts Progressive Growing of Generative Adversarial Networks(PG-GANs) as an implementation model to generate high-resolution images and to enhance variation of the generated images, and applied it to fashion image data. PG-GANs allows the generator and discriminator to progressively learn at the same time, continuously adding new layers from low-resolution images to result high-resolution images. We also proposed a Mini-batch Discrimination method to increase the diversity of generated data, and proposed a Sliced Wasserstein Distance(SWD) evaluation method instead of the existing MS-SSIM to evaluate the GAN model.

• Journal of Korea Society of Industrial Information Systems
• /
• v.19 no.2
• /
• pp.85-91
• /
• 2014

The objective of the research is to automatically recognize railway objects from MLS data in which 9 key objects including terrain, track, bed, vegetation, platform, barrier, posts, attachments, powerlines are targeted. The proposed method can be divided into two main sub-steps. First, multi-scale contextual features are extracted to take the advantage of characterizing objects of interest from different geometric levels such as point, line, volumetric and vertical profile. Second, by considering contextual interactions amongst object labels, a contextual classifier is utilized to make a prediction with local coherence. In here, the Conditional Random Field (CRF) is used to incorporate the object context. By maximizing the object label agreement in the local neighborhood, CRF model could compensate the local inconsistency prediction resulting from other local classifiers. The performance of proposed method was evaluated based on the analysis of commission and omission error and shows promising results for the practical use.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.2C
• /
• pp.65-77
• /
• 2011

With the rapid expansion of application fields of vector data model such as CAD design drawing and GIS digital map, the security technique for vector data model has been issued. This paper presents the vector data hashing for the authentication and copy protection of vector data model. The proposed hashing groups polylines in main layers of a vector data model and generates the group coefficients by the line curve curvatures of the first and second type of all poly lines. Then we calculate the feature coefficients by projecting the group coefficients onto the random pattern and generate finally the binary hash from the binarization of the feature coefficients. From experimental results using a number of CAD drawings and GIS digital maps, we verified that the proposed hashing has the robustness against various attacks and the uniqueness and security by the random key.

• Informatization Policy
• /
• v.25 no.4
• /
• pp.108-120
• /
• 2018

With the rise of the Fourth Industrial Revolution and I-Korea 4.0, both of which pursue strategies for industrial innovation and for the solution to social problems, the real estate industry needs to change in order to make effective use of available space in smart environments. The implementation of smart spaces is a promising solution for this. The smart space is defined as a good use of space, whether it be a home, office, or retail store, within a smart environment. To enhance the use of smart spaces, efficient decision-making and well-timed and accurate interaction are required. This paper proposes a smart space based on platform which takes advantage of emerging technologies for the efficient storage, processing, analysis, and utilization of big data. The platform is composed of six layers - collection, transfer, storage, service, application, and management - and offers three service frameworks: activity-based, market-based, and policy-based. Based on these smart space services, decision-makers, consumers, clients, and social network participants can make better decisions, respond more quickly, exhibit greater innovation, and develop stronger competitive advantages.