• Journal of Korean Society of Steel Construction
• /
• v.17 no.1 s.74
• /
• pp.1-11
• /
• 2005

A bi-level damage detection algorithm that utilizes the dynamic responses of the structure as input and neural network (NN) as pattern classifier is presented. Signal anomaly index (SAI) is proposed to express the amount of changes in the shape of frequency response functions (FRF) or strain frequency response function (SFRF). SAI is calculated using the acceleration and dynamic strain responses acquired from intact and damaged states of the structure. In a bi-level damage identification algorithm, the presence of damage is first identified from the magnitude of the SAI value, then the location of the damage is identified using the pattern recognition capability of NN. The proposed algorithm is applied to an experimental model bridge to demonstrate the feasibility of the algorithm. Numerically simulated signals are used for training the NN, and experimentally-acquired signals are used to test the NN. The results of this example application suggest that the SAI-based pattern recognition approach may be applied to the structural health monitoring system for a real bridge.

• The KIPS Transactions:PartB
• /
• v.10B no.4
• /
• pp.443-450
• /
• 2003

In this paper, we proposed the novel hierarchical algorithm for the recognition of English calling cards that processes multiresolution images of calling cards hierarchically to extract individual characters and recognizes the extracted characters by using the enhanced neural network method. The hierarchical recognition algorithm generates multiresolution images of calling cards, and each processing step in the algorithm selects and processes the image with suitable resolution for lower processing overhead and improved output. That is, first, the image of 1/3 times resolution, to which the horizontal smearing method is applied, is used to extract the areas including only characters from the calling card image, and next, by applying the vertical smearing and the contour tracking masking, the image of a half time resolution is used to extract individual characters from the character string areas. Lastly, the original image is used in the recognition step, because the image includes the morphological information of characters accurately. And for the recognition of characters with diverse font types and various sizes, the enhanced RBF network that improves the middle layer based on the ART1 was proposed and applied. The results of experiments on a large number of calling card images showed that the proposed algorithm is greatly improved in the performance of character extraction and recognition compared with the traditional recognition algorithms.

• Journal of the Korean Geotechnical Society
• /
• v.37 no.11
• /
• pp.61-69
• /
• 2021

In this study, scale down model bridge piers were fabricated and non-destructive experiments conducted with an impact load to determine scours in the ground adjacent to the bridge piers using the natural frequency of the bridge piers. Three scale-model bridge piers with different heights were fabricated, and they penetrated the ground at a depth of 0.35 m. The scours around the bridge piers were simulated as a side scour and foundation scour. The experiments were conducted in 13 steps, in which scouring around the model bridge piers was performed in 0.05 m excavation units. To derive the natural frequency, the impact load was measured with three accelerometers attached to the model bridge piers. The impact load was applied with an impact hammer, and the top of the model bridge pier was struck perpendicularly to the bridge axis. The natural frequency according to the scour progress was calculated with a fast Fourier transform. The results demonstrated that the natural frequency of each bridge pier tended to decrease with scour progress. The natural frequency also decreased with increasing pier height. With scour progress, a side scour occurred at 70% or higher of the initial natural frequency, and a foundation scour occurred at less than 70%.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.48 no.6
• /
• pp.411-418
• /
• 2020

In this paper, approximation and optimization methods are proposed for the structural performance of the composite fan blade. Using these methods, we perform the optimal design of the stacking sequence to maximize stiffnesses without changing the mass and the geometric shape of the composite fan blade. In this study, the lamination parameters are introduced to reduce the design variables and space. From the characteristics of lamination parameters, we generate response surface model having a high fitness value. Considering the requirements of the optimal stacking sequence, the multi-objective optimization problem is formulated. We apply the two-step optimization method that combines gradient-based method and genetic algorithm for efficient search of an optimal solution. Finally, the finite element analysis results of the initial and the optimized model are compared to validate the approximation and optimization methods based on the lamination parameters.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.18 no.2
• /
• pp.117-123
• /
• 2018

In this study, we suggest a fast image reconstruction scheme using Poisson equation from image gradient domain. In this approach, using the Poisson equation, a guided vector field is created by employing source and target images within a selected region at the first step. Next, the guided vector is used in generating the result image. We analyze the problem of reconstructing a two-dimensional function that approximates a set of desired gradients and a data term. The joined data and gradients are able to work like modifying the image gradients while staying close to the original image. Starting with this formulation, we have a screened Poisson equation known in physics. This equation leads to an efficient solution to the problem in FFT domain. It represents the spatial filters that solve the two-dimensional screened Poisson model and shows gradient scaling to be a well-defined sharpen filter that generalizes Laplace sharpening. We demonstrate the results using a discrete cosine transformation based this Poisson model.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.12 no.4
• /
• pp.11-21
• /
• 2013

The conventional real-time signal control strategy for oversaturated situation generally requires a number of detectors at the intersection in order to identify the queue length at each approach. Also, existing strategy for the spillback has limited effect due to the temporal spillback control which only reduce the green split at the approach. In this study, a signal control system utilizing the imagery information from ITS roadside equipment is developed for operation of oversaturated intersections. The strategy calculates the saturation ratio based on the queue length extracted from ITS RSE, and designs the signal control variables according to the saturation ratio. The signal control strategy is divided into two phases: oversaturated and supersaturated. In oversaturated conditions, timing plan for main approach is optimized by the queue length. In oversaturated conditions where spillback might occur, the signal timing is designed in order to avoid the spillback. To increase field adaptability, the strategy is designed to adjust the split length, all-red-time, and clearance time, and keep the major signal control variables intact. The result of the simulation shows that in oversaturated conditions, the improvement is similar to the real-time signal control system. In case of, oversaturated conditions, however, the effect of the strategy is superior to that of a real-time system.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.15 no.3
• /
• pp.171-177
• /
• 2013

The purpose of this study is to find a scheme to scale down the KMA (Korea Meteorological Administration) digital precipitation maps to the grid cell resolution comparable to the rural landscape scale in Korea. As a result, we suggest two steps procedure called RATER (Radar Assisted Topography and Elevation Revision) based on both radar echo data and a mountain precipitation model. In this scheme, the radar reflection intensity at the constant altitude of 1.5 km is applied first to the KMA local analysis and prediction system (KLAPS) 5 km grid cell to obtain 1 km resolution. For the second step the elevation and topography effect on the basis of 270 m digital elevation model (DEM) which represented by the Parameter-elevation Regressions on Independent Slopes Model (PRISM) is applied to the 1 km resolution data to produce the 270 m precipitation map. An experimental watershed with about $50km^2$ catchment area was selected for evaluating this scheme and automated rain gauges were deployed to 13 locations with the various elevations and slope aspects. 19 cases with 1 mm or more precipitation per day were collected from January to May in 2013 and the corresponding KLAPS daily precipitation data were treated with the second step procedure. For the first step, the 24-hour integrated radar echo data were applied to the KLAPS daily precipitation to produce the 1 km resolution data across the watershed. Estimated precipitation at each 1 km grid cell was then regarded as the real world precipitation observed at the center location of the grid cell in order to derive the elevation regressions in the PRISM step. We produced the digital precipitation maps for all the 19 cases by using RATER and extracted the grid cell values corresponding to 13 points from the maps to compare with the observed data. For the cases of 10 mm or more observed precipitation, significant improvement was found in the estimated precipitation at all 13 sites with RATER, compared with the untreated KLAPS 5 km data. Especially, reduction in RMSE was 35% on 30 mm or more observed precipitation.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.24 no.2
• /
• pp.217-226
• /
• 2006

Extraction and tracking objects are fundamental and important steps of the digital image processing and computer vision. Many algorithms about extracting and tracking objects have been developed. In this research, a method is suggested for tracking a moving object using a pair of CCD cameras and calculating the coordinate of the moving object. A 1/100 miniature of soccer field was made to apply the developed algorithms. After candidates were selected from the acquired images using the RGB value of a moving object (soccer ball), the object was extracted using its size (MBR size) among the candidates. And then, image coordinates of a moving object are obtained. The real-time position of a moving object is tracked in the boundary of the expected motion, which is determined by centering the moving object. The 3D position of a moving object can be obtained by conducting the relative orientation, absolute orientation, and space intersection of a pair of the CCD camera image.

• Journal of Intelligence and Information Systems
• /
• v.16 no.2
• /
• pp.95-108
• /
• 2010

In this paper, we present a hybrid neural network model for dynamic hand gesture recognition. The model consists of two modules, feature extraction module and pattern classification module. We first propose a modified CNN(convolutional Neural Network) a pattern recognition model for the feature extraction module. Then we introduce a weighted fuzzy min-max(WFMM) neural network for the pattern classification module. The data representation proposed in this research is a spatiotemporal template which is based on the motion information of the target object. To minimize the influence caused by the spatial and temporal variation of the feature points, we extend the receptive field of the CNN model to a three-dimensional structure. We discuss the learning capability of the WFMM neural networks in which the weight concept is added to represent the frequency factor in training pattern set. The model can overcome the performance degradation which may be caused by the hyperbox contraction process of conventional FMM neural networks. From the experimental results of human action recognition and dynamic hand gesture recognition for remote-control electric home appliances, the validity of the proposed models is discussed.

• Journal of the Korean GEO-environmental Society
• /
• v.15 no.12
• /
• pp.117-128
• /
• 2014

Generally, when constructing a tunnel close to existing structures, the tunnel must be built at a constant distance from the structures that is more than width of tunnel to minimize the impact of interference between an existing structures and new tunnel. Spacing of these closed tunnels should be designed considering soil state, size of tunnel and reinforcement method. Particularly when the ground is soft, a care should be taken with the tunnel plans because the closer the tunnel is to the existing structures, the greater the deformation becomes. As methods of reviewing the effect of cavities on the stability of a tunnel, field measurement, numerical analysis and scaled model test can be considered. In the methods, the scaled model test can reproduce the engineering characteristics of a rock in a field condition and the shape of structures using the scale factor even not all conditions cannot be considered. In this study, when construction of a tunnel close to existing structures, the method and considering factors of the scaled model test were studied to predict the actual tunnel behavior in planning stage. Furthermore, model test results were compared with the numerical analysis results for verifying the proposed model test procedure. Also, practical results were derived to verify the stability of a tunnel vis-a-vis cavities through the scaled model test, which assumed spacing distances of 0.25 D, 0.50 D, and 1.00 D between the cavities and tunnel as well as the network state distribution. The spacing distances of 1.0 D is evaluated as the critical distance by the results of model test and numerical analysis.