• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.2
• /
• pp.93-98
• /
• 2016

For accurate precipitation forecasts the choice of weather factors and prediction method is very important. Recently, machine learning has been widely used for forecasting precipitation, and artificial neural network, one of machine learning techniques, showed good performance. In this paper, we suggest a new method for forecasting precipitation using DBN, one of deep learning techniques. DBN has an advantage that initial weights are set by unsupervised learning, so this compensates for the defects of artificial neural networks. We used past precipitation, temperature, and the parameters of the sun and moon's motion as features for forecasting precipitation. The dataset consists of observation data which had been measured for 40 years from AWS in Seoul. Experiments were based on 8-fold cross validation. As a result of estimation, we got probabilities of test dataset, so threshold was used for the decision of precipitation. CSI and Bias were used for indicating the precision of precipitation. Our experimental results showed that DBN performed better than MLP.

• Restorative Dentistry and Endodontics
• /
• v.27 no.4
• /
• pp.421-431
• /
• 2002

C-shaped canal configuration is very difficult to treat because that clues about preoperative canal anatomy cannot be ascertained from clinical crown morphology and limited information can be derived from radiographic examination. This study was done to get more informations about the root and canal configuration of C-shape root by 3-dimensionally reconstructing for the purpose of enhancing success rate of endodontic treatment. 30 mandibular molars with C-shaped root were selected. Six photo images from occlusal, apical, mesial, distal, buccal, lingual directions and radiographic view were taken as preoperative ones to compare them with 3-D image. After crown reduction to the level of 1-2mm over pulpal floor was performed, teeth were stored in 5.25% sodium hypochlorite solution for the removal of pulp tissue and debris. They were cleaned under running water, allowed to bench dry and embedded in a self-curing resin. This resin block was serially ground with a microtome (Accutom-50, Struers, Denmark) and the image of each level was recorded by digital camera (FinePix S1-pro, Fuji Co., Japan). The thickness of each section was 0.25mm. Photographs of serial sections through all root canal were digitized using Adobe Photoshop 5.0 and then minimum thickness of open and closed sites were measured (open site is the surface containing occluso-apical groove closed site is oppsite). After dizitization using 3-D Doctor (Able software Corp, USA). 3D reconstruction of the outer surface of tooth and the inner surface of pulp space was made. Canal classsification of C-shaped roots was performed from this 3-D reconstructed image. The results were as follows : 1. Most C-shape rooted teeth showed lingual groove (28/30). 2 According to Vertuccis' calssification, type I, II, III, IV, VII were observed. but also new canal types suck as 2-3-2, 1-2-3-2. 2-3-2-1, 2-3-2-3 were shown. 3 There was little difference in minimum thickness on coronal and apical portions, but open site were thinner than closed site on mid portion. Conclusively, 3D reconstruction method could make the exact configurations of C-shape root possible to be visualized and analyzed from multi-directions. Data from minimum thickness recommend cleaning and shaping be more carefully done on dangerous mid portion.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.45 no.6
• /
• pp.117-122
• /
• 2008

Although the pulse diagnosis is the one of the most important diagnostic process to traditional medical doctors, there is no proper communication tool between experts and trainees. In this paper, we have developed a indentation training system which consists of a hardware measuring indent pressure on artificial arm quantitatively and a software providing a indentation training program. The hardware for measurement of indent pressure profile includes 3 load cells embedded in the artificial arm, signal amplification part and digitization part, NI-USB 6009 with 200Hz sampling rate. For setting up a relationship table between weights and output voltages, 8 standard weights were used. To evaluate this hardware, 3 oriental medical specialists were involved and their indent pressure profile were recorded three times respectively. From these, it was found that pulse diagnosis process could be divided into 3 periods and the maximum load were $500g{\cdot}f$ approximately while doctors perform a pulse diagnosis. The indentation training program was implemented with LabView and designed to monitor the differences between the pressure profile of a expert and that of a trainee so to offer some visual feedback to the trainee. Also, this program could provide the trends of training performances. With this developed system, the education of pulse diagnosis is expected to be more quantitative and effective.

• Journal of Intelligence and Information Systems
• /
• v.8 no.1
• /
• pp.141-157
• /
• 2002

extensible Markup Language (XML) is a new markup language for data exchange on the Internet. In this paper, we propose a language extensible Rule Markup Language (XRML) which is an extension of XML. The implicit rules embedded in the Web pages should be identifiable, interchangeable with structured rule format, and finally accessible by various applications. It is possible to realize by using XRML. In this light, Web based Knowledge Management Systems (KMS) can be integrated with rule-based expert systems. To meet this end, we propose the six design criteria: Expressional Completeness, Relevance Linkability, Polymorphous Consistency, Applicative Universality, Knowledge Integrability and Interoperability. Furthermore, we propose three components such as RIML (Rule Identification Markup Language), RSML (Rule Structure Markup Language) and RTML (Rule Triggering Markup Language), and the Document Type Definition DTD). We have designed the XRML version 0.5 as illustrated above, and developed its prototype named Form/XRML which is an automated form processing for disbursement of the research fund in the Korea Advanced Institute of Science and Technology (KAISI). Since XRML allows both human and software agent to use the rules, there is huge application potential. We expect that XRML can contribute to the progress of Semantic Web platforms making knowledge management and e-commerce more intelligent. Since there are many emerging research groups and vendors who investigate this issue, it will not take long to see XRML commercial products. Matured XRML applications may change the way of designing information and knowledge systems in the near future.

• Korean Journal of Remote Sensing
• /
• v.30 no.1
• /
• pp.149-160
• /
• 2014

Recently studies on application development and utilization using sensors and devices embedded in smartphones have flourished at home and abroad. This study aimed to analyze the accuracy of the images of smartphone to determine three-dimension position of close objects prior to the development of photogrammetric system applying smartphone and evaluate the feasibility to use. First of all, camera calibration was conducted on autofocus and infinite focus. Regarding camera calibration distortion model with balance system and unbalance system was used for the decision of lens distortion coefficient, the results of calibration on 16 types of projects showed that all cases were in RMS error by less than 1 mm from bundle adjustment. Also in terms of autofocus and infinite focus on S and S2 model, the pattern of distorted curve was almost the same, so it could be judged that change in distortion pattern according to focus mode is very little. The result comparison according to autofocus and infinite focus and the result comparison according to a software used for multi-image processing showed that all cases were in standard deviation less than ${\pm}3$ mm. It is judged that there is little result difference between focus mode and determination of three-dimension position by distortion model. Lastly the checkpoint performance by total station was fixed as most probable value and the checkpoint performance determined by each project was fixed as observed value to calculate statistics on residual of individual methods. The result showed that all projects had relatively large errors in the direction of Y, the direction of object distance compared to the direction of X and Z. Like above, in terms of accuracy for determination of three-dimension position for a close object, the feasibility to use smartphone camera would be enough.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.40 no.9
• /
• pp.597-604
• /
• 2016

A parallel algorithm of bi-conjugate gradient method was developed based on CUDA for parallel computation of the incompressible Navier-Stokes equations. The governing equations were discretized using splitting P2P1 finite element method. Asymmetric stenotic flow problem was solved to validate the proposed algorithm, and then the parallel performance of the GPU was examined by measuring the elapsed times. Further, the GPU performance for sparse matrix-vector multiplication was also investigated with a matrix of fluid-structure interaction problem. A kernel was generated to simultaneously compute the inner product of each row of sparse matrix and a vector. In addition, the kernel was optimized to improve the performance by using both parallel reduction and memory coalescing. In the kernel construction, the effect of warp on the parallel performance of the present CUDA was also examined. The present GPU computation was more than 7 times faster than the single CPU by double precision.

• Journal of Korea Multimedia Society
• /
• v.7 no.3
• /
• pp.368-376
• /
• 2004

Currently, the package tools and software development kits of Microsoft's Windows Media Technology provides users an environment to write, distribute and play multimedia contents. However, it is very difficult for common users to create a media streaming content using the Windows Media technology, because of its technical hardship. Therefore, we developed an easy-to-use authoring tool for multimedia streaming content which can be used in the Windows Media Technology environment. Our authoring tool is based on the web page templates by which HTML codes can be generated automatically. Using the existing streaming media authoring tool provided by the Microsoft, users have to code the temporal markers and the event script commands directly to the ASF(Advanced Systems Format) files or ASX (ASF Stream Redirector) format files. We developed an easy-to-use interface for insetting temporal markers and the event script commands to the ASF files by simple mouse clicking and dragging to minimize the effort of users for creating multimedia streaming content. In addition, we implemented an ASX file creation wizard that allows users to create ASX format file without coding ASX tags. We also implemented various web page templates where the ASF files and the ASX files can be embedded by just clicking the corresponding templates. The target web pages can be automatically generated from those templates and previewed directly on the web browser at the time of creation.

• Journal of KIISE
• /
• v.43 no.6
• /
• pp.710-717
• /
• 2016

Recently, several research studies on things abstraction framework have been proposed in order to implement the multi-service Internet of Things (IoT) system, where various IoT services share the thing devices. Distributed things abstraction has an IoT service duplication problem, which aggravates power consumption of mobile devices and network traffic. On the other hand, cloud server-centric things abstraction cannot cover real-time interactions due to long network delay. Fog server-centric things abstraction has limits in insufficient IoT interfaces. In this paper, we propose Smart Fog which is a fog server-centric things abstraction framework to resolve the problems of the existing things abstraction frameworks. Smart Fog consists of software modules to operate the Smart Gateway and three interfaces. Smart Fog is implemented based on IoTivity framework and OIC standard. We construct a smart home prototype on an embedded board Odroid-XU3 using Smart Fog. We evaluate the network performance and energy efficiency of Smart Fog. The experimental results indicate that the Smart Fog shows short network latency, which can perform real-time interaction. The results also show that the proposed framework has reduction in the network traffic of 74% and power consumption of 21% in mobile device, compared to distributed things abstraction.

• Journal of KIISE:Software and Applications
• /
• v.37 no.1
• /
• pp.29-38
• /
• 2010

We propose an interactive projection technique to display details of a large image in a high resolution and brightness by tracking a portable projector. A closed-loop based tracking method is presented to update the projected image while a user changes the position of the detail area by moving the portable projector. A marker is embedded in the large image to indicate the position to be occupied by the detail image projected by the portable projector. The marker is extracted in sequential images acquired by a camera attached to the portable projector. The marker position in the large display image is updated under a constraint that the center positions of marker and camera frame coincide in every camera frame. The image and projective transformation for warping are calculated using the marker position and shape in the camera frame. The marker's four corner points are determined by a four-step segmentation process which consists of camera image preprocessing based on HSI, edge extraction by Hough transformation, quadrangle test, and cross-ratio test. The interactive projection system implemented by the proposed method performs at about 24fps. In the user study, the overall feedback about the system usability was very high.

• The KIPS Transactions:PartA
• /
• v.16A no.3
• /
• pp.159-168
• /
• 2009

In this paper, we present an efficient way of implementing OpenGL ES 1.1 standard for the environments with hardware-supported OpenGL API, such as desktop PCs. Although OpenGL ES was started from the existing OpenGL features, it becomes a new three-dimensional graphics library customized for embedded systems through introducing fixed-point arithmetic operations, buffer management with fixed-point data type supports, completely new texture mapping functionalities and others. Currently, it is the official three dimensional graphics library for Google Android, Apple iPhone, PlayStation3, etc. In this paper, we achieved improvements on the arithmetic operations for the fixed-point number representation, which is the most characteristic data type for OpenGL ES. For the conversion of fixed-point data types to the floating-point number representations for the underlying OpenGL, we show the way of efficient conversion processes even with satisfying OpenGL ES standard requirements. We also introduced a simple memory management scheme to mange the converted data for the buffer containing fixed-point numbers. In the case of texture processing, the requirements in both standards are quite different and thus we used completely new software-implementations. Our final implementation result of OpenGL ES library provides all of over than 200 functions in OpenGL ES 1.1 standard and completely passed its conformance test, to show its compliance with the standard. From the efficiency viewpoint, we measured its execution times for several OpenGL ES-specific application programs and achieved at most 33.147 times improvements, to become the fastest one among the OpenGL ES implementations in the same category.