• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.12
• /
• pp.1835-1845
• /
• 2021

To accurately identify black ice and warn the drivers of information in advance so they can control speed and take preventive measures. In this paper, we propose a lightweight black ice detection network based on infrared road images. A black ice recognition network model based on CNN transfer learning has been developed. Additionally, to further improve the accuracy of black ice recognition, an enhanced lightweight network based on MobileNetV2 has been developed. To reduce the amount of calculation, linear bottlenecks and inverse residuals was used, and four bottleneck groups were used. At the same time, to improve the recognition rate of the model, each bottleneck group was connected to a 3×3 convolutional layer to enhance regional feature extraction and increase the number of feature maps. Finally, a black ice recognition experiment was performed on the constructed infrared road black ice dataset. The network model proposed in this paper had an accurate recognition rate of 99.07% for black ice.

• Korean Journal of Remote Sensing
• /
• v.38 no.5_3
• /
• pp.819-834
• /
• 2022

Artificial Intelligence (AI) techniques have been effectively used for image classification, object detection, and image segmentation. Along with the recent advancement of computing power, deep learning models can build deeper and thicker networks and achieve better performance by creating more appropriate feature maps based on effective activation functions and optimizer algorithms. This review paper examined technical and academic trends of Convolutional Neural Network (CNN) and Transformer models that are emerging techniques in remote sensing and suggested their utilization strategies and development directions. A timely supply of satellite images and real-time processing for deep learning to cope with disaster monitoring will be required for future work. In addition, a big data platform dedicated to satellite images should be developed and integrated with drone and Closed-circuit Television (CCTV) images.

• The Korean Journal of Archival Studies
• /
• no.33
• /
• pp.113-153
• /
• 2012

Since the government of Republic of Korea was established in 1948, a period that made the biggest difference on National Records Management System was from 1999 when was enacted. Especially, it was the period of President Roh's five-year tenure called Participation Government (2003-2008). This paper illustrates distinct phenomena of Records Management System Policy during Participation Government. Three major agents of the system are President Roh, the Presidential Secretariat, and Archives Management Bureau at the National Archives of Korea. They sometimes competed with themselves for initiatives of policy, but they used to cooperate with each other and have brought about innovations on records management. The first distinctive characteristic of Participation Government (below PG)'s records management is that it implemented governance actively. That is, it tried to listen carefully to all opinions of interest organizations related to records management and enacted laws based on those. The PG not only listened to civic groups, but also created two professional groups called Records Management Innovation Expert Committee and Innovation Decentralization Assessment Committee. Those two groups enacted . Another remarkable feature is a nomination of records management specialists at public institutions. In 2005, PG created Archival Research Positions among research public officials and appointed experts in the field of Archival Research History at central department. With the process, the government tried to provide public records management system and to improve specialty of records management. Since then, records management specialists were employed not only at local governments but also at private archival institutions. It has allowed of entering a new phase in employing records management professionals. The Participation Government also legislated (completely revised) . It led to a beginning of developing records management in Republic of Korea. was revised thoroughly for the e-Government period and was established as a foundation for managing presidential records. An establishing process of a country's records management system describes the degree of democratic development of society. Following governments should supplement PG's shortcomings and carry out 'New Governance Records Management System'. Principal subjects of records management system should include not only a government but also civic groups, local governments, small businesses, and academic professionals. The object of records management also needs to be democratic by recording not only the plans and enforcements of a task but also influences and results of a task. The way of archiving ought to be discussed by all related principals.

• The Journal of Society for e-Business Studies
• /
• v.11 no.3
• /
• pp.35-51
• /
• 2006

Information visualization plays an important role to provide a conceptual comprehension of the information and should be adjusted and featured to reflect the characteristics of the domain to which it is applied. Product searching in e-Commerce is not an exception. In this paper, we present visualization techniques that are specified for effectively browsing the results of product searching. We considered visualization techniques including MapNet and Cluster Map and modified them to reflect the characteristics of the product domain. We consider two types of search queries: category searching and product property searching, and make it possible to include more detailed product semantics in their search criteria. We also provide a query refinement mechanism so that even users with rack of understanding the products may rewrite their queries for better results.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.26 no.5
• /
• pp.85-91
• /
• 1989

In this paper, we propose a neural net system for speech recognition, which is composed of two neural networks. Firstly the self-supervised BP(Back Propagation) network generates the distributed concept corresponding to the activity pattern in the hidden units. And then the self-organizing neural network forms a concept map which directly displays the similarity relations between concepts. By doing the above, the difficulty in learning the conventional BP network is solved and the weak side of BP falling into a pattern matcher is gone, while the strong point of generating the various internal representations is used. And we have obtained the concept map which is more orderly than the Kohonen's SOFM. The proposed neural net system needs not any special preprocessing and has a self-learning ability.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.4A
• /
• pp.345-355
• /
• 2002

This paper describes a new method of directional filter-based analysis for fingerprint enhancement. Fingerprint cages can be represented by direction field of regular structure of ridge patterns. The dominant directional component of ridge plays a very important role in pre-processing steps of fingerprint image analysis such as ridge's linking and noise removal for minutiae extraction. A directional filter bank analyzes input image into directional subband images and synthesizes them to the perfectly reconstructed image. In this paper, a new fingerprint enhancement algorithm based on a directional filter bank is proposed. The algorithm decomposes the fingerprint image into subband images in the analysis stage, accomplishes an enhance procedure by processing subband images in the enhance stage and synthesizes them to the enhanced image in the synthesis stage.

• Journal of Korea Society of Industrial Information Systems
• /
• v.11 no.3
• /
• pp.34-39
• /
• 2006

This paper introduces a network security situation awareness tool using a traffic pattern map which facilitates recognizing a current network status by extracting and analyzing predetermined traffic features and displaying an abnormal or harmful traffic which deteriorates network performance. The traffic pattern-map consists of $26{\times}26$ intersections, on which the occupancy rate of the port having maximum occupancy is displayed as a bar graph. In general, in case of the Internet worm, the source address section on the traffic pattern map is activated. In case of DDoS the destination address section is activated.

• Journal of Korea Game Society
• /
• v.18 no.3
• /
• pp.113-122
• /
• 2018

In this paper, we will discuss planning and production of VR contents for effective racing drone steering training. In the real world, direct racing drone training training is obviously a space and economic limit. We propose through the production of VR-based drone infiltration action game as a solution to these limitations and try to improve effective drone handling ability. Especially, we designed the virtual reality contents in a similar environment by using the similarity of FPV goggles and HMD based VR, and designed virtual drones, maps and obstacles considering the characteristics of the structures used in racing drone and drone racing. Also, we tried to increase the user's immersion and presence by increasing difficulty level of obstacles.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2002.04a
• /
• pp.365-368
• /
• 2002

본 논문에서는 SANtopiaVM의 데이터 가용성 및 성능을 고려한 에러 핸들링 정책을 제안한다. 제안하는 에러 핸들링 정책은 RAID1과 RAID5에 대한 정책으로 추분하며 에러 발생시의 복구를 위하여 RAID1에서는 FBB(Failed Block Bitmap)라는 비트맵을 추가하여 데이터의 일관성을 유지하고 디스크의 복구 시에는 저기 연산에 대한 부하 분산 및 복구 비용을 감소시키는 쓰기 연산을 제안함으로서 에러 핸들링 시에 입출력 비용을 줄인다. RAID5에서는 추가적인 여분 디스크를 사용한 Sparing Disk 기법을 제안함으로서 디스크 에러 모드에서도 정상 모드의 성능에 근접하는 입출력 성능을 보장하며 빠른 디스크 복구를 지원한다. 제안하는 에러 핸들링 정책은 SANtopiaVM RAID의 오류 발생시, 시스템 성능의 급격한 저하를 완화할 수 있으며 에러로부터 빠른 복구를 지원함으로서 데이터에 대한 효율적인 고 가용성의 특징을 제공한다.

• Journal of Institute of Control, Robotics and Systems
• /
• v.17 no.5
• /
• pp.436-443
• /
• 2011

Recognition of structures in urban environments is a fundamental ability for unmanned ground vehicles. In this paper we propose the geometrical featured voxel which has not only 3-D coordinates but also the type of geometrical properties of point cloud. Instead of dealing with a huge amount of point cloud collected by range sensors in urban, the proposed voxel can efficiently represent and save 3-D urban structures without loss of geometrical properties. We also provide an urban structure classification algorithm by using the proposed voxel and machine learning techniques. The proposed method enables to recognize urban environments around unmanned ground vehicles quickly. In order to evaluate an ability of the proposed map representation and the urban structure classification algorithm, our vehicle equipped with the sensor system collected range data and pose data in campus and experimental results have been shown in this paper.