• Title/Summary/Keyword: Automatic validation

Search Result 186, Processing Time 0.03 seconds

A Study on the Automatic Detection of Railroad Power Lines Using LiDAR Data and RANSAC Algorithm (LiDAR 데이터와 RANSAC 알고리즘을 이용한 철도 전력선 자동탐지에 관한 연구)

  • Jeon, Wang Gyu;Choi, Byoung Gil
    • Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
    • /
    • v.31 no.4
    • /
    • pp.331-339
    • /
    • 2013
  • LiDAR has been one of the widely used and important technologies for 3D modeling of ground surface and objects because of its ability to provide dense and accurate range measurement. The objective of this research is to develop a method for automatic detection and modeling of railroad power lines using high density LiDAR data and RANSAC algorithms. For detecting railroad power lines, multi-echoes properties of laser data and shape knowledge of railroad power lines were employed. Cuboid analysis for detecting seed line segments, tracking lines, connecting and labeling are the main processes. For modeling railroad power lines, iterative RANSAC and least square adjustment were carried out to estimate the lines parameters. The validation of the result is very challenging due to the difficulties in determining the actual references on the ground surface. Standard deviations of 8cm and 5cm for x-y and z coordinates, respectively are satisfactory outcomes. In case of completeness, the result of visual inspection shows that all the lines are detected and modeled well as compare with the original point clouds. The overall processes are fully automated and the methods manage any state of railroad wires efficiently.

Sensitivity Analysis of the High-Resolution WISE-WRF Model with the Use of Surface Roughness Length in Seoul Metropolitan Areas (서울지역의 고해상도 WISE-WRF 모델의 지표면 거칠기 길이 개선에 따른 민감도 분석)

  • Jee, Joon-Bum;Jang, Min;Yi, Chaeyeon;Zo, Il-Sung;Kim, Bu-Yo;Park, Moon-Soo;Choi, Young-Jean
    • Atmosphere
    • /
    • v.26 no.1
    • /
    • pp.111-126
    • /
    • 2016
  • In the numerical weather model, surface properties can be defined by various parameters such as terrain height, landuse, surface albedo, soil moisture, surface emissivity, roughness length and so on. And these parameters need to be improved in the Seoul metropolitan area that established high-rise and complex buildings by urbanization at a recent time. The surface roughness length map is developed from digital elevation model (DEM) and it is implemented to the high-resolution numerical weather (WISE-WRF) model. Simulated results from WISE-WRF model are analyzed the relationship between meteorological variables to changes in the surface roughness length. Friction speed and wind speed are improved with various surface roughness in urban, these variables affected to temperature and relative humidity and hence the surface roughness length will affect to the precipitation and Planetary Boundary Layer (PBL) height. When surface variables by the WISE-WRF model are validated with Automatic Weather System (AWS) observations, NEW experiment is able to simulate more accurate than ORG experiment in temperature and wind speed. Especially, wind speed is overestimated over $2.5m\;s^{-1}$ on some AWS stations in Seoul and surrounding area but it improved with positive correlation and Root Mean Square Error (RMSE) below $2.5m\;s^{-1}$ in whole area. There are close relationship between surface roughness length and wind speed, and the change of surface variables lead to the change of location and duration of precipitation. As a result, the accuracy of WISE-WRF model is improved with the new surface roughness length retrieved from DEM, and its surface roughness length is important role in the high-resolution WISE-WRF model. By the way, the result in this study need various validation from retrieved the surface roughness length to numerical weather model simulations with observation data.

Automated Vehicle Research by Recognizing Maneuvering Modes using LSTM Model (LSTM 모델 기반 주행 모드 인식을 통한 자율 주행에 관한 연구)

  • Kim, Eunhui;Oh, Alice
    • The Journal of The Korea Institute of Intelligent Transport Systems
    • /
    • v.16 no.4
    • /
    • pp.153-163
    • /
    • 2017
  • This research is based on the previous research that personally preferred safe distance, rotating angle and speed are differentiated. Thus, we use machine learning model for recognizing maneuvering modes trained per personal or per similar driving pattern groups, and we evaluate automatic driving according to maneuvering modes. By utilizing driving knowledge, we subdivided 8 kinds of longitudinal modes and 4 kinds of lateral modes, and by combining the longitudinal and lateral modes, we build 21 kinds of maneuvering modes. we train the labeled data set per time stamp through RNN, LSTM and Bi-LSTM models by the trips of drivers, which are supervised deep learning models, and evaluate the maneuvering modes of automatic driving for the test data set. The evaluation dataset is aggregated of living trips of 3,000 populations by VTTI in USA for 3 years and we use 1500 trips of 22 people and training, validation and test dataset ratio is 80%, 10% and 10%, respectively. For recognizing longitudinal 8 kinds of maneuvering modes, RNN achieves better accuracy compared to LSTM, Bi-LSTM. However, Bi-LSTM improves the accuracy in recognizing 21 kinds of longitudinal and lateral maneuvering modes in comparison with RNN and LSTM as 1.54% and 0.47%, respectively.

The Analysis and Implication of Link Integrity of Korean Public Institution Website from a Web Usability Perspective (웹 사용성 관점에서 공공기관 웹 사이트의 링크 유효성 분석 및 개선 과제)

  • Moon, Hyun Ju;Kim, Suk Il
    • 재활복지
    • /
    • v.17 no.4
    • /
    • pp.291-309
    • /
    • 2013
  • Web usability is utilized as a standard for constructing and managing website to be useful by people of all abilities and disabilities. Link integrity is one of main criteria for web usability evaluation. Links that are found to be broken or pointing to irrelevant information lead gradual degradation of link integrity over time. This study has conducted investigation for link integrity of Korean public institution website. Among 49 website that are accessible by an automatic link checker software, 91.8% of the website has more than one broken links. Also average number of broken links is 50.5 that is 0.33% of the number of links. 29 websites consist of internal broken links. And 22 website do not provide any information on the missing page. Broken links enlower website confidence by reducing web usability as well as they prohibit web search engine not to provide information on the website. Therefore, web developers and managers need to be aware of importance of link integrity and prohibit web usability degradation by performing link validation as a web management task or periodically.

Validation on the algorithm of estimation of collision risk among ships based on AIS data of actual ships' collision accident (선박충돌사고 AIS 데이터 기반 선박 충돌위험도 추정 알고리즘 검증에 관한 연구)

  • Son, Nam-Sun;Kim, Sun-Young
    • Proceedings of the Korean Institute of Navigation and Port Research Conference
    • /
    • 2010.10a
    • /
    • pp.180-181
    • /
    • 2010
  • An estimation algorithm of collision risk among multiple ships has been developed in order to reduce human error and prevent collision accidents. The algorithm is designed to calculate the collision risk among ships based on Fuzzy theory by using AIS data as traffic information. In this paper, to validate the algorithm, the AIS data of actual collision accident, which occurred between a product carrier and a cargo carrier in Busan harbor in 2009 are collected. The replay simulation is carried out on the actual AIS data and the collision risk is calculated in real time. In this paper, the features of the estimation algorithm of collision risk and the results of replay simulation based on AIS data of actual collision accident are discussed.

  • PDF

Study on Fault Diagnosis Method of Train Communication Network applied to the prototype Korean High Speed Train

  • Cho, Chang-Hee;Park, Min-Kook;Kwon, Soon-Man;Kim, Yong-Ju;Kim, Sung-Shin
    • 제어로봇시스템학회:학술대회논문집
    • /
    • 2003.10a
    • /
    • pp.2169-2173
    • /
    • 2003
  • The development project of Korean High Speed Train (KHST) was started in 1996. As a national research project, the KHST project aims for a development of the next generation prototype train that has a maximum speed of 350 km/h. The development process of prototype KHST including 7 vehicles was completed last year and currently the prototype train is on its way of test running over the test track with gradually increased speed. The prototype KHST uses the real time network called TCN (Train Communication Network) for exchanging information between various onboard control equipments. After 10 years of development and modification period, TCN was confirmed as international standard (IEC61375-1) for the electrical railway equipment train bus. In the prototype KHST, all major control devices are connected by TCN and exchange their information. Such devices include SCU (Supervisory Control Unit), ATC (Automatic Train Control), TCU (Traction Control Unit), and so forth. For each device that sends and receives data using TCN, a device has to find out whether TCN is in normal or failure state before its data exchange. And also a device must have a proper method of data validation that was received in a normal TCN state. This is a one of the major important factors for devices using network. Some misleading information can lead the entire system to a catastrophic condition. This paper briefly explains how TCN was implemented in the prototype KHST train, and also shows what kind of the fault diagnosis method was adopted for a fail safe operation of TCN system

  • PDF

Automatic Construction of SHACL Schemas for RDF Knowledge Graphs Generated by Direct Mappings

  • Choi, Ji-Woong
    • Journal of the Korea Society of Computer and Information
    • /
    • v.25 no.10
    • /
    • pp.23-34
    • /
    • 2020
  • In this paper, we proposes a method to automatically construct SHACL schemas for RDF knowledge graphs(KGs) generated by Direct Mapping(DM). DM and SHACL are all W3C recommendations. DM consists of rules to transform the data in an RDB into an RDF graph. SHACL is a language to describe and validate the structure of RDF graphs. The proposed method automatically translates the integrity constraints as well as the structure information in an RDB schema into SHACL. Thus, our SHACL schemas are able to check integrity instead of RDBMSs. This is a consideration to assure database consistency even when RDBs are served as virtual RDF KGs. We tested our results on 24 DM test cases, published by W3C. It was shown that they are effective in describing and validating RDF KGs.

Quality Level Classification of ECG Measured using Non-Constraint Approach (무구속적 방법으로 측정된 심전도의 신뢰도 판별)

  • Kim, Y.J.;Heo, J.;Park, K.S.;Kim, S.
    • Journal of Biomedical Engineering Research
    • /
    • v.37 no.5
    • /
    • pp.161-167
    • /
    • 2016
  • Recent technological advances in sensor fabrication and bio-signal processing enabled non-constraint and non-intrusive measurement of human bio-signals. Especially, non-constraint measurement of ECG makes it available to estimate various human health parameters such as heart rate. Additionally, non-constraint ECG measurement of wheelchair user provides real-time health parameter information for emergency response. For accurate emergency response with low false alarm rate, it is necessary to discriminate quality levels of ECG measured using non-constraint approach. Health parameters acquired from low quality ECG results in inaccurate information. Thus, in this study, a machine learning based approach for three-class classification of ECG quality level is suggested. Three sensors are embedded in the back seat, chest belt, and handle of automatic wheelchair. For the two sensors embedded in back seat and chest belt, capacitively coupled electrodes were used. The accuracy of quality level classification was estimated using Monte Carlo cross validation. The proposed approach demonstrated accuracy of 94.01%, 95.57%, and 96.94% for each channel of three sensors. Furthermore, the implemented algorithm enables classification of user posture by detection of contacted electrodes. The accuracy for posture estimation was 94.57%. The proposed algorithm will contribute to non-constraint and robust estimation of health parameter of wheelchair users.

Hyperparameter Search for Facies Classification with Bayesian Optimization (베이지안 최적화를 이용한 암상 분류 모델의 하이퍼 파라미터 탐색)

  • Choi, Yonguk;Yoon, Daeung;Choi, Junhwan;Byun, Joongmoo
    • Geophysics and Geophysical Exploration
    • /
    • v.23 no.3
    • /
    • pp.157-167
    • /
    • 2020
  • With the recent advancement of computer hardware and the contribution of open source libraries to facilitate access to artificial intelligence technology, the use of machine learning (ML) and deep learning (DL) technologies in various fields of exploration geophysics has increased. In addition, ML researchers have developed complex algorithms to improve the inference accuracy of various tasks such as image, video, voice, and natural language processing, and now they are expanding their interests into the field of automatic machine learning (AutoML). AutoML can be divided into three areas: feature engineering, architecture search, and hyperparameter search. Among them, this paper focuses on hyperparamter search with Bayesian optimization, and applies it to the problem of facies classification using seismic data and well logs. The effectiveness of the Bayesian optimization technique has been demonstrated using Vincent field data by comparing with the results of the random search technique.

A comparison of deep-learning models to the forecast of the daily solar flare occurrence using various solar images

  • Shin, Seulki;Moon, Yong-Jae;Chu, Hyoungseok
    • The Bulletin of The Korean Astronomical Society
    • /
    • v.42 no.2
    • /
    • pp.61.1-61.1
    • /
    • 2017
  • As the application of deep-learning methods has been succeeded in various fields, they have a high potential to be applied to space weather forecasting. Convolutional neural network, one of deep learning methods, is specialized in image recognition. In this study, we apply the AlexNet architecture, which is a winner of Imagenet Large Scale Virtual Recognition Challenge (ILSVRC) 2012, to the forecast of daily solar flare occurrence using the MatConvNet software of MATLAB. Our input images are SOHO/MDI, EIT $195{\AA}$, and $304{\AA}$ from January 1996 to December 2010, and output ones are yes or no of flare occurrence. We consider other input images which consist of last two images and their difference image. We select training dataset from Jan 1996 to Dec 2000 and from Jan 2003 to Dec 2008. Testing dataset is chosen from Jan 2001 to Dec 2002 and from Jan 2009 to Dec 2010 in order to consider the solar cycle effect. In training dataset, we randomly select one fifth of training data for validation dataset to avoid the over-fitting problem. Our model successfully forecasts the flare occurrence with about 0.90 probability of detection (POD) for common flares (C-, M-, and X-class). While POD of major flares (M- and X-class) forecasting is 0.96, false alarm rate (FAR) also scores relatively high(0.60). We also present several statistical parameters such as critical success index (CSI) and true skill statistics (TSS). All statistical parameters do not strongly depend on the number of input data sets. Our model can immediately be applied to automatic forecasting service when image data are available.

  • PDF