• Geomechanics and Engineering
• /
• v.37 no.5
• /
• pp.467-474
• /
• 2024

Monitoring and managing the condition of underground utilities is crucial for ground stability. This study aims to determine whether images obtained using ground penetrating radar (GPR) accurately reflect the characteristics of buried pipelines through image analysis. The investigation focuses on pipelines made from different materials, namely concrete and steel, with concrete pipes tested under various diameters to assess detectability under differing conditions. A total of 400 images are acquired at locations with pipelines, and for comparison, an additional 100 data points are collected from areas without pipelines. The study employs GPR at frequencies of 200 MHz and 600 MHz, and image analysis is performed using machine learning-based convolutional neural network (CNN) techniques. The analysis results demonstrate high classification reliability based on the training data, especially in distinguishing between pipes of the same material but of different diameters. The findings suggest that the integration of GPR and CNN algorithms can offer satisfactory performance in exploring the ground's interior characteristics.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 1999.10a
• /
• pp.365-373
• /
• 1999

Recently, numerous studies have demonstrated that artificial intelligence such as neural networks can be an alternative methodology for classification problems to which traditional statistical methods have long been applied. In building neural network model, the selection of independent and dependent variables should be approached with great care and should be treated as a model construction process. Irrespective of the efficiency of a learning procedure in terms of convergence, generalization and stability, the ultimate performance of the estimator will depend on the relevance of the selected input variables and the quality of the data used. Approaches developed in statistical methods such as correlation analysis and stepwise selection method are often very useful. These methods, however, may not be the optimal ones for the development of neural network models. In this paper, we propose a genetic algorithms approach to find an optimal or near optimal input variables for neural network modeling. The proposed approach is demonstrated by applications to bankruptcy prediction modeling. Our experimental results show that this approach increases overall classification accuracy rate significantly.

• Water Engineering Research
• /
• v.4 no.2
• /
• pp.59-68
• /
• 2003

The applicabilities and validities of two methodologies fur the prediction of THM (trihalomethane) formation in a water pipeline system were proposed and discussed. One is the multiple regression technique and the other is an artificial neural network technique. There are many factors which influence water quality, especially THMs formations in water pipeline systems. In this study, the prediction models of THM formation in water pipeline systems are developed based on the independent variables proposed by American Water Works Association(AWWA). Multiple linear/nonlinear regression models are estimated and three layer feed-forward artificial neural networks have been used to predict the THM formation in a water pipeline system. Input parameters of the models consist of organic compounds measured in water pipeline systems such as TOC, DOC and UV254. Also, the reaction time to each measuring site along pipeline is used as input parameter calculated by a hydraulic analysis. Using these variables as model parameters, four models are developed. And the predicted results from the four developed models are compared statistically to the measured THMs data set. It is shown that the artificial neural network approaches are much superior to the conventional regression approaches and that the developed models by neural network can be used more efficiently and reproduce more accurately the THMs formation in water pipeline systems, than the conventional regression methods proposed by AWWA.

• Journal of Sensor Science and Technology
• /
• v.28 no.3
• /
• pp.182-186
• /
• 2019

When a camera is employed for 3D sensing, accurate camera calibration is vital as it is a prerequisite for the subsequent steps of the sensing process. Camera calibration is usually performed by complex mathematical modeling and geometric analysis. On the other contrary, data learning using an artificial neural network can establish a transformation relation between the 3D space and the 2D camera image without explicit camera modeling. However, a neural network requires a large amount of accurate data for its learning. A significantly large amount of time and work using a precise system setup is needed to collect extensive data accurately in practice. In this study, we propose a two-step neural calibration method that is effective when only a small amount of learning data is available. In the first step, the camera projection transformation matrix is determined using the limited available data. In the second step, the transformation matrix is used for generating a large amount of synthetic data, and the neural network is trained using the generated data. Results of simulation study have shown that the proposed method as valid and effective.

• Journal of the Korean Society for Precision Engineering
• /
• v.14 no.11
• /
• pp.135-146
• /
• 1997

This paper describes fault diagnosis method in complex system with hierachical structure similar to human body structure. Complex system is divided into unit, item and component. For diagnosing this hierarchical complex system, it is necessary to implement special neural network. Fault diagnosis system can forecast faults in a system and decide from current machine state signal information. Comparing with other diagnosis system for single fault, the developed system deals with multiple fault diagnosis comprising Hierarchical Neural Network(HNN). HNN consists of four level neural network, first level for item fault symptom classification, second level for item fault diagnosis, third level for component symptom classification, forth level for component fault diagnosis. UNIX IPC(Inter Process Communication) is used for implementing HNN wiht multitasking and message transfer between processes in SUN workstation with X-Windows(Motif). We tested HNN at four units, seven items per unit, seven components per item in a complex system. Each one neural newtork operate as a separate process in HNN. The message queue take charge of information exdhange and cooperation between each neural network.

• The Journal of the Acoustical Society of Korea
• /
• v.37 no.2
• /
• pp.118-128
• /
• 2018

In this paper, we have studied how to search an underwater object by learning the image generated by the side scan sonar in the convolution neural network. In the method of human side analysis of the side scan image or the image, the convolution neural network algorithm can enhance the efficiency of the analysis. The image data of the side scan sonar used in the experiment is the public data of NSWC (Naval Surface Warfare Center) and consists of four kinds of synthetic underwater objects. The convolutional neural network algorithm is based on Faster R-CNN (Region based Convolutional Neural Networks) learning based on region of interest and the details of the neural network are self-organized to fit the data we have. The results of the study were compared with a precision-recall curve, and we investigated the applicability of underwater object detection in convolution neural networks by examining the effect of change of region of interest assigned to sonar image data on detection performance.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.05a
• /
• pp.473-476
• /
• 2003

Nowaday, Internet is doing the role of a large distributed information service tenter and various information and database servers managing it are in distributed network environment. However, the we have several difficulties in deciding the server to disposal input data depending on data properties. In this paper, we designed server mining mechanism and Intellectual data mining system architecture for the best efficiently dealing with input data pattern by using neural network among the various data in distributed environment. As a result, the new input data pattern could be operated after deciding the destination server according to dynamic binding method implemented by neural network. This mechanism can be applied Datawarehous, telecommunication and load pattern analysis, population census analysis and medical data analysis.

• Computers and Concrete
• /
• v.20 no.1
• /
• pp.31-38
• /
• 2017

In the present study an Artificial Neural Network (ANN) was used to predict the compressive strength of self-compacting concrete. The data developed experimentally for self-compacting concrete and the data sets of a total of 99 concrete samples were used in this work. ANN's are considered as nonlinear statistical data modeling tools where complex relationships between inputs and outputs are modeled or patterns are found. In the present ANN model, eight input parameters are used to predict the compressive strength of self-compacting of concrete. These include varying amounts of cement, coarse aggregate, fine aggregate, fly ash, fiber, water, super plasticizer (SP), viscosity modifying admixture (VMA) while the single output parameter is the compressive strength of concrete. The importance of different input parameters for predicting the strengths at various ages using neural network was discussed in the study. There is a perfect correlation between the experimental and prediction of the compressive strength of SCC based on ANN with very low root mean square errors. Also, the efficiency of ANN model is better compared to the multivariable regression analysis (MRA). Hence it can be concluded that the ANN model has more potential compared to MRA model in developing an optimum mix proportion for predicting the compressive strength of concrete without much loss of material and time.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 2000.11b
• /
• pp.270-278
• /
• 2000

A machine vision system using charge coupled device(CCD) camera for the weed detection in a radish farm was developed. Shape features were analyzed with the binary images obtained from color images of radish and weeds. Aspect, Elongation and PTB were selected as significant variables for discriminant models using the STEPDISC option. The selected variables were used in the DISCRIM procedure to compute a discriminant function for classifying images into one of the two classes. Using discriminant analysis, the successful recognition rate was 92% for radish and 98% for weeds. To recognize radish and weeds more effectively than the discriminant analysis, an artificial neural network(ANN) was used. The developed ANN model distinguished the radish from the weeds with 100%. The performance of ANNs was improved to prevent overfitting and to generalize well using a regularization method. The successful recognition rate in the farms was 93.3% for radish and 93.8% for weeds. As a whole, the machine vision system using CCD camera with the artificial neural network was useful to detect weeds in the radish farms.

• The Transactions of the Korea Information Processing Society
• /
• v.2 no.3
• /
• pp.417-424
• /
• 1995

This paper describes MFT(Mean Field Theory) neural network with continuous with continuous variables is applied to quantification analysis problem. A quantification analysis problem, one of the important problems in statistics, is NP complete and arises in the optimal location of objects in the design space according to the given similarities only. This paper presents a MFT neural network with continuous variables for the quantification problem. Starting with reformulation of the quantification problem to the penalty problem, this paper propose a "one-variable stochastic simulated annealing(one-variable SSA)" based on the mean field approximation. This makes it possible to evaluate of the spin average faster than real value calculating in the MFT neural network with continuous variables. Consequently, some experimental results show the feasibility of this approach to overcome the difficulties to evaluate the spin average value expressed by the integral in such models.ch models.