• ETRI Journal
• /
• v.39 no.1
• /
• pp.97-107
• /
• 2017

A fast and accurate building-level visual place recognition method built on an image-retrieval scheme using street-view images is proposed. Reference images generated from street-view images usually depict multiple buildings and confusing regions, such as roads, sky, and vehicles, which degrades retrieval accuracy and causes matching ambiguity. The proposed practical database refinement method uses informative reference image and keypoint selection. For database refinement, the method uses a spatial layout of the buildings in the reference image, specifically a building-identification mask image, which is obtained from a prebuilt three-dimensional model of the site. A global-positioning-system-aware retrieval structure is incorporated in it. To evaluate the method, we constructed a dataset over an area of $0.26km^2$. It was comprised of 38,700 reference images and corresponding building-identification mask images. The proposed method removed 25% of the database images using informative reference image selection. It achieved 85.6% recall of the top five candidates in 1.25 s of full processing. The method thus achieved high accuracy at a low computational complexity.

• Proceedings of the Korean Society of Agricultural Engineers Conference
• /
• 1999.10c
• /
• pp.354-358
• /
• 1999

In practicing the refinement project of agriuctural production basys, many researches have been done until now, but they are not reused because management is not easily accomplished. We analysed and designed the Components of refinement project by the object -orientation technique, and presented a method of accumulation and management of object's data using relational database. In result, management of new data is easy and reusibility are increased compared to structural analysis technique.

• Health Policy and Management
• /
• v.14 no.1
• /
• pp.121-147
• /
• 2004

Since the pilot program for a DRG-based prospective payment system was introduced in 1997, the performance of KDRGs has been one of hotly debated issues. The objectives of this study are to refine the classification algorithm of the KDRGs and to assess the improvement achieved by the refinement. The U.S. Medicare DRGs version 17.0 and the Australian Refined DRGs version 4.1 were reviewed to identify areas of possible impro-vement. Refined changes in the classification and result of date analyses were submitted to a panel of 48 physicians for their reviews and suggestions. The refinement was evaluated by the variance reduction in resource utilization achieved by the KDRG The database of 2,182,168 claims submitted to the Health Insurance Review Agency during 2002 was used for evaluation. As the result of the refinement, three new MDCs were introduced and the number of ADEGs increased from 332 to 674. Various age splits and two to four levels of severity classification for secondary diagnoses were introduced as well. A total of 1,817 groups were defined in the refined KDRGs. The variance reduction for charges of all patients increased from 48.2％ to 53.6％ by the refinement, and from 65.6％ to 73.1％ for non-outlier patients. The r-square for length of stays of all patients was increased from 28.3％ to 32.6％, and from 40.4％ to 44.9％ for non-outlier patients. These results indicated a significant improvement in the classification accuracy of the KDRG system.

• The Journal of Information Technology and Database
• /
• v.4 no.2
• /
• pp.33-45
• /
• 1998

Database summarization is the procedure to obtain generalized and representative descriptions expressing the content of a large amount of database at a glance. We present a top-down summary refinement procedure to discover database summaries. The procedure exploits attribute concept hierarchies that represent ISA relationships among domain concepts. It begins with the most generalized summary and proceeds to find more specialized ones by stepwise refinements. This top-down paradigm reveals at least two important advantages compared to the previous bottom-up methods. Firstly, it provides a natural way of reflecting the user's own discovery preference interactively. Secondly, it does not produce too large intermediate result that makes it hard for the bottom-up approach to be applied in practical environment. The proposed procedure can also be easily extended for distributed databases. Information content measure of a database summary is derived in order to identify more informative summaries among the discovered results.

• Journal of Korea Multimedia Society
• /
• v.18 no.12
• /
• pp.1453-1461
• /
• 2015

Since digital images and videos are rapidly increasing in the internet with the spread of mobile computers and smartphones, research on image retrieval has gained tremendous momentum. Color, shape, and texture are major features used in image retrieval. Especially, color information has been widely used in image retrieval, because it is robust in translation, rotation, and a small change of camera view. This paper proposes a new method for histogram refinement based on local color difference. Firstly, the proposed method converts a RGB color image into a HSV color image. Secondly, it reduces the size of color space from 256³ to 32. It classifies pixels in the 32-color image into three groups according to the color difference between a central pixel and its neighbors in a 3x3 local region. Finally, it makes a color difference vector(CDV) representing three refined color histograms, then image retrieval is performed by the CDV matching. The experimental results using public image database show that the proposed method has higher retrieval accuracy than other conventional ones. They also show that the proposed method can be effectively applied to search low resolution images such as thumbnail images.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.11 no.1
• /
• pp.1-7
• /
• 2011

Most of expert systems use the text-oriented knowledge bases. However, knowledge management using the knowledge bases is considered as a huge burden to the knowledge workers because it includes some troublesome works. It includes chasing and/or checking activities on Consistency, Redundancy, Circulation, and Refinement of the knowledge. In those cases, we consider that they could reduce the burdens by using relational database management systems-based knowledge management infrastructure and convert the knowledge into one of easy forms human can understand. Furthermore they could concentrate on the knowledge itself with the support of the systems. To meet the expectations, in this study, we have tried to develop a general-purposed knowledge conversion tool for expert systems. Especially, this study is focused on the knowledge conversions among text-oriented knowledge base, relational database knowledge base, and decision tree.

• Journal of KIISE:Databases
• /
• v.30 no.3
• /
• pp.286-295
• /
• 2003

The shape of query windows for spatial selection queries is a rectangle in many cases. However, it can be issued for spatial selection queries with not only rectangular query widow, but also polygonal query window. Moreover, as the applications like GIS can manage much more spatial data, they can support the more various applications. Therefore it is valuable for considering about the query processing method suitable for not only rectangle query window, but also general polygonal one. It is the general state-of-the-art approach to use the plane- sweep technique as the computation algorithm in the refinement step as the spatial join queries do. However, from the observation on the characteristics of spatial data and query windows, we can find in many cases that the shape of query window is much simpler than that of spatial data. From these observations, we suggest a new refinement process approach which is suitable for this situation. Our experiments show that, if the number of vertices composing the query window is less than about 20, the new approach we suggest is superior to the state-of-the-art approach by about 20% in general cases.

• The Journal of Korean Medicine
• /
• v.34 no.1
• /
• pp.170-178
• /
• 2013

Objectives: Because TKM-related statistics are scattered in year books, white papers, or web services, it is not easy to use. This study aimed to build a TKM statistical database offering one-stop service. Methods: We collected TKM statistics from year books, white papers, and web services using TKM related keywords which came from the advanced study. Then we removed, merged, or split the statistics looking through data and meta-data of statistics to build a database. Results: First, we classified the statistics into 6 categories and found that the service category was 303, product category 37 and so on. Second, we analyzed the source of statistics and found that the Ministry of Health & Welfare and the Korea statistics were the strongest sources. Third, we analyzed the years of statistics and found that the oldest was 1962, 35% were within the past 3 years, and only 14% were over 10 years old. Finally, we found that the 415 statistics were composed of 1,120 items and the average was 2.70. Conclusions: Through this study we found that more sophisticated TKM statistics must be produced and developed specially in the education, research, and administration categories. We hope that the effort of TKM societies is available to eliminate the lack of statistical producer's understanding about TKM.

• The KIPS Transactions:PartD
• /
• v.10D no.4
• /
• pp.591-602
• /
• 2003

Delay and discontinuance phenomenon of service are cause by sudden increase of the network communication amount and the quantity consumed of resources when Internet users are driven excessively to a conventional single large database sewer. To solve these problems, spatial database cluster consisted of several single nodes on high-speed network to offer high-performance is risen. But, research about spatial join operation that can reduce the performance of whole system in case process at single node is not achieved. So, in this paper, we propose efficient parallel spatial join processing method in a spatial database cluster system that uses data partitions and replications method that considers the characteristics of space data. Since proposed method does not need the creation step and the assignment step of tasks, and does not occur additional message transmission between cluster nodes that appear in existent parallel spatial join method, it shows performance improvement of 23% than the conventional parallel R-tree spatial join for a shared-nothing architecture about expensive spatial join queries. Also, It can minimize the response time to user because it removes redundant refinement operation at each cluster node.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.2
• /
• pp.135-144
• /
• 2003

A method for the automatic segmentation of speech signals is described. The method is dedicated to the construction of a large database for a Text-To-Speech (TTS) synthesis system. The main issue of the work involves the refinement of an initial estimation of phone boundaries which are provided by an alignment, based on a Hidden Market Model(HMM). Multi-layer perceptron (MLP) was used as a phone boundary detector. To increase the performance of segmentation, a technique which individually trains an MLP according to phonetic transition is proposed. The optimum partitioning of the entire phonetic transition space is constructed from the standpoint of minimizing the overall deviation from hand labelling positions. With single speaker stimuli, the experimental results showed that more than 95% of all phone boundaries have a boundary deviation from the reference position smaller than 20 ms, and the refinement of the boundaries reduces the root mean square error by about 25%.