• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.6
• /
• pp.774-784
• /
• 2021

The object detection and recognition process is a very important task in the field of computer vision, and related research is actively being conducted. However, in the actual object recognition process, the recognition accuracy is often degraded due to the resolution mismatch between the training image data and the test image data. To solve this problem, in this paper, we designed and developed an integrated object recognition and super-resolution framework by proposing an image super-resolution technique to improve object recognition accuracy. In detail, 11,231 license plate training images were built by ourselves through web-crawling and artificial-data-generation, and the image super-resolution artificial neural network was trained by defining an objective function to be robust to the image flip. To verify the performance of the proposed algorithm, we experimented with the trained image super-resolution and recognition on 1,999 test images, and it was confirmed that the proposed super-resolution technique has the effect of improving the accuracy of character recognition.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.119-127
• /
• 2022

Most of the construction works are conducted outdoors, so the construction workers are affected by weather conditions such as temperature, humidity, and wind velocity which can be evaluated the thermal comfort as environmental factors. In our previous researches, it was found that construction accidents are usually occurred in the discomfort ranges. The safety management, therefore, should be planned in consideration of the thermal comfort and measured by a specialized simulation tool. However, it is very complex, time-consuming, and difficult to model. To address this issue, this study is aimed to develop a framework of a prediction model for improving the prediction accuracy about outdoor thermal comfort considering environmental factors using machine learning algorithms with hyperparameter tuning. This study is done in four steps: i) Establishment of database, ii) Selection of variables to develop prediction model, iii) Development of prediction model; iv) Conducting of hyperparameter tuning. The tree type algorithm is used to develop the prediction model. The results of this study are as follows. First, considering three variables related to environmental factor, the prediction accuracy was 85.74%. Second, the prediction accuracy was 86.55% when considering four environmental factors. Third, after conducting hyperparameter tuning, the prediction accuracy was increased up to 87.28%. This study has several contributions. First, using this prediction model, the thermal comfort can be calculated easily and quickly. Second, using this prediction model, the safety management can be utilized to manage the construction accident considering weather conditions.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.8
• /
• pp.1606-1616
• /
• 2011

The blood pressure measuring equipment, which is being supplied and used most widely by being recognized convenience and accuracy now generally, is oscillometric blood pressure monitor. However, a change in blood pressure is basically influenced by diverse elements such as each individual's physiological status and physical condition. Thus, the measurement of blood pressure, which used single element called oscillation in blood pressure of being conveyed to cuff, is not considered on physiological elements such as cardiovascular system status and blood vessel stiffness index, and on external elements, thereby being quite in error. Accordingly, this study detected diverse bio-signals and body informations in each individual as the measurement subject such as ECG, PPG, and Korotkoff Sound in order to enhance convenience and accuracy of measuring blood pressure in the complex measurement equipment, thereby having extracted regression method for compensation in error of oscillometric blood pressure measurement on the wrist, and having improved accuracy of measuring blood pressure. To verify a method of improving accuracy, the blood pressure value in each of SBP, DBP, MAP was acquired through 4-stage experimental procedure targeting totally 51 subjects. Prior to experiment, the subjects were divided into two groups such as the experimental group for extracting regression method and the control group for verifying regression method. Its error was analyzed by comparing the reference blood pressure value, which was obtained through the auscultatory method, and the oscillometric blood pressure value on the wrist. To reduce the detected error, the blood pressure compensation regression method was calculated through multiple linear regression analysis on elements of blood pressure, individual body information, PTT, HR, K-Sound PSD change. Verification was carried out on improving significance and accuracy by applying the regression method to the data of control group. In the experimental results, as a result of confirming error on the reference blood pressure value in SBP, DBP, and MAP, which were acquired through applying regression method, the results of $-0.47{\pm}7.45$ mmHg, $-0.23{\pm}7.13$ mmHg, $0.06{\pm}6.39$ mmHg could be obtained. This is not only the numerical value of satisfying the sphygmomanometer reference of AAMI, but also shows the lower result than the numerical value in SBP : $-2.5{\pm}12.2$ mmHg, DBP : $-7.5{\pm}8.4$ mmHg, which is the mean error in the experimental results of Brram's research for verifying accuracy of Omron RX-M, which shows relatively high accuracy among wrist sphygmomanometers. Thus, the blood pressure compensation could be confirmed to be made within significant level.

• Proceedings of the Korea Contents Association Conference
• /
• 2014.06a
• /
• pp.487-488
• /
• 2014

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.4 no.1
• /
• pp.79-92
• /
• 2000

We investigate numerical properties of Newton's algorithm for improving an eigenpair of a real matrix A using only fixed precision arithmetic. We show that under natural assumptions it produces an eigenpair of a componentwise small relative perturbation of the data matrix A.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2002.08a
• /
• pp.122-126
• /
• 2002

Measuring color precisely requires the use of a colorimeter whose spectral response is as close as possible to the CIE response. Some techniques such as the NIST 4-color method can improve the precision. but under given conditions, the results can be worse.

• Journal of the Korean Institute of Navigation
• /
• v.24 no.3
• /
• pp.157-165
• /
• 2000

When the GPS system was come to the operation, the U.S Coast Guard initiated development of differential GPS system based on a marine radiobeacon. It has rapidly been spread out to many countries. DGPS equals to P(Y) code in a positioning accuracy. USA has recently determined to expand the coverage of DGPS to inland in order to install a nationwide DGPS chain. Korea is under processing for improving the DGPS as a nationwide positioning system. Before expanding the service area inside Korea, we need to verify the relation between the field strength and DGPS accuracy for the service area. The Japanese DGPS data is received in the southern Part of the Korean peninsula. The Korean DGPS was not a complete system, so we selected the Japanese DGPS data as a model for the study. This paper investigate accuracy and reliability characteristics of RBN/DGPS over the effective service area. Through the experimental and simulation study, we obtained the reliable and stable positioning accuracy in the southern part of the Korean peninsula. In addition, the characteristics of RBN/DGPS were examined in the land over the effective coverage from Japan. The results would be a basic reference to research the RBN/NDGPS in Korea.

• Journal of Korean Society for Geospatial Information Science
• /
• v.14 no.4 s.38
• /
• pp.3-10
• /
• 2006

This study proposes a method to measure the positional accuracy of the implemented GIS polygon data. Also, it aims to present a possibility to analyze the occurrence types of positional errors by improving the measuring methods of positional accuracy based on the existing individual methods and by linking individual methods. As a result of the actual application of the methodology to the test area, it was possible to measure the positional accuracy in target test areas and to analyze the occurrence causes (types) of positional errors through each index linking (linking methodologies). Also, research results allowed confirming the applicability of the methodology. However, complementary research for each standard numerical value is recommended in order to ensure the validity of methodology.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.3
• /
• pp.1348-1375
• /
• 2018

Support Vector Machine (SVM) is a well-known machine learning classification algorithm, which has been widely applied to many data mining problems, with good accuracy. However, SVM classification speed decreases with increase in dataset size. Some applications, like video surveillance and intrusion detection, requires a classifier to be trained very quickly, and on large datasets. Hence, this paper introduces two filter-based instance selection techniques for optimizing SVM training speed. Fast classification is often achieved at the expense of classification accuracy, and some applications, such as phishing and spam email classifiers, are very sensitive to slight drop in classification accuracy. Hence, this paper also introduces two wrapper-based instance selection techniques for improving SVM predictive accuracy and training speed. The wrapper and filter based techniques are inspired by Cuckoo Search Algorithm and Bat Algorithm. The proposed techniques are validated on three popular e-fraud types: credit card fraud, spam email and phishing email. In addition, the proposed techniques are validated on 20 other datasets provided by UCI data repository. Moreover, statistical analysis is performed and experimental results reveals that the filter-based and wrapper-based techniques significantly improved SVM classification speed. Also, results reveal that the wrapper-based techniques improved SVM predictive accuracy in most cases.

• Journal of Korean Society on Water Environment
• /
• v.32 no.6
• /
• pp.582-588
• /
• 2016

The hydrological simulation program FORTRAN (HSPF) is a comprehensive watershed model that employs the hydraulic function table (FTABLE) (depth-area-volume-flow relationship) to represent the geometric and hydraulic properties of water bodies. The hydraulic representation of the HSPF model mainly depends on the accuracy of the FTABLES. These hydraulic representations determine the response time of water quality state variables and also control the scour, deposition, and transport of sediments in the water body. In general, FTABLES are automatically generated based on reach information such as mean depth, mean width, length, and slope along with a set of standard assumptions about the geometry and hydraulics of the channel, so these FTABLES are unable to accurately describe the geometry and hydraulic behavior of rivers and reservoirs. In order to compensate the weakness of HSPF for hydraulic modeling, we generated alternate method to improve the accuracy of FTABLES for rivers, using the surveyed cross sections and rating curves. The alternative method is based on the hydraulics simulated by HEC-RAS using the surveyed cross sections and rating curves, and it could significantly improve the accuracy of FTABLES. Although the alternate FTABLE greatly improved the hydraulic accuracy of the HSPF model, it had little effect on the hydrological simulation.