• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2008.04a
• /
• pp.65-70
• /
• 2008

The capacity spectrum method (CSM) is a deterministic seismic analysis approach wherein the expected seismic response of a structure is established as the intersection of the demand and capacity curves. Recently, there are a few studies about a probabilistic CSM where uncertainties in design factors such as material properties, loads, and ground motion are being considered. However, researches show that soil-structure interaction also affects the seismic responses of structures. Thus, their uncertainties should also be taken into account. Therefore, this paper presents a probabilistic approach of using the CSM for seismic analysis considering uncertainties in soil properties. For application, a reinforced concrete bridge column structure is employed as a test model. Considering the randomness of the various design parameters, the structure's probability of failure is obtained. Monte Carlo importance sampling is used as the tool to assess the structure's reliability when subjected to earthquakes. In this study, probabilistic CSM with and without consideration of soil uncertainties are compared and analyzed. Results show that the analysis considering soil structure interaction yields to a greater probability of failure, and thus can lead to a more conservative structural design.

• Journal of Korea Multimedia Society
• /
• v.6 no.7
• /
• pp.1302-1311
• /
• 2003

This paper presents a more stable method of multiresolution editing for a triangular mesh. The basic idea of our paper is to embed an editing area of a mesh onto a 2D region and to produce 3D surfaces which interpolate the editing-information. In this paper, we adopt the extended convex combination approach based on the shape-preserving parameterization for the embedding, which guarantees no self-intersection on the 2D embedded mesh. That is, the result of the embedding is stable. Moreover, we adopt the multi-level B-spline approach to generate the surface containing all of 3D editing-information, which can make us control the editing area in several levels. Hence, this method supports interactive editing and thus can produce intuitive editing results.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 1993.10a
• /
• pp.1254-1263
• /
• 1993

Representing knowledge uncertainty , aggregating evidence confidences , and propagation uncertainties are three key elements that effect the ability of a rule-based expert system to represent domains with uncertainty . Fuzzy set theory provide a good mathematical tool for representing the vagueness associated with a variable when , as the condition of a rule , it only partially corresponds to the input data. However, the aggregation of ANDed and Ored confidences is not as simple as the intersection and union operators defined for fuzzy set membership. There is, in fact, a certain degree of compensation that occurs when an expert aggregates confidences associated with compound evidence . Further, expert often consider individual evidences to be varying importance , or weight , in their support for a conclusion. This paper presents a flexible approach for evaluating evidence and conclusion confidences. Evidences may be represented as fuzzy or nonfuzzy variables with as associat d degree of certainty . different weight can also be associated degree of certainty. Different weights can also be assigned to the individual condition in determining the confidence of compound evidence . Conclusion confidence is calculated using a modified approach combining the evidence confidence and a rule strength. The techniques developed offer a flexible framework for representing knowledge and propagating uncertainties. This framework has the potention to reflect human aggregation of uncertain information more accurately than simple minimum and maximum operator do.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.15 no.3
• /
• pp.119-127
• /
• 2020

The detection and recognition of abnormal driving becomes crucial for achieving safety in Intelligent Transportation Systems (ITS). This paper presents a feature extraction method based on spectral data to train a neural network model for driving behavior recognition. The proposed method uses a two stage signal processing approach to derive time-saving and efficient feature vectors. For the first stage, the feature vector set is obtained by calculating variances from each frequency bin containing the power spectrum data. The feature set is further reduced in the second stage where an intersection method is used to select more significant features that are finally applied for training a neural network model. A stream of live signals are fed to the trained model which recognizes the abnormal driving behaviors. The driving behaviors considered in this study are weaving, sudden braking and normal driving. The effectiveness of the proposed method is demonstrated by comparing with existing methods, which are Particle Swarm Optimization (PSO) and Convolution Neural Network (CNN). The experiments show that the proposed approach achieves satisfactory results with less computational complexity.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.221-235
• /
• 2022

Manual inspection of steel box girders on long span bridges is time-consuming and labor-intensive. The quality of inspection relies on the subjective judgements of the inspectors. This study proposes an automated approach to detect and segment cracks in high-resolution images. An end-to-end cascaded framework is proposed to first detect the existence of cracks using a deep convolutional neural network (CNN) and then segment the crack using a modified U-Net encoder-decoder architecture. A Naïve Bayes data fusion scheme is proposed to reduce the false positives and false negatives effectively. To generate the binary crack mask, first, the original images are divided into 448 × 448 overlapping image patches where these image patches are classified as cracks versus non-cracks using a deep CNN. Next, a modified U-Net is trained from scratch using only the crack patches for segmentation. A customized loss function that consists of binary cross entropy loss and the Dice loss is introduced to enhance the segmentation performance. Additionally, a Naïve Bayes fusion strategy is employed to integrate the crack score maps from different overlapping crack patches and to decide whether a pixel is crack or not. Comprehensive experiments have demonstrated that the proposed approach achieves an 81.71% mean intersection over union (mIoU) score across 5 different training/test splits, which is 7.29% higher than the baseline reference implemented with the original U-Net.

• Journal of Korean Society of Transportation
• /
• v.31 no.1
• /
• pp.77-86
• /
• 2013

This paper focuses on the development of analytical models for estimating the changes in saturation flow rates (SFR) at the stop-lines of a signalized intersection due to the existence of nearby work-zones, and thereby calculating the prevailing capacity values for specific lane groups. Major changes were incorporated in the logics of previous models and significant revisions have been made to secure the accuracy and simplicity. Furthermore, much attention was paid to model validation by making comparisons to both extensive simulation results and empirical data from various sites. It was found that SFRs are highly sensitive to the location of work-zones, the distance to each work-zone from the stop-line of a concerned approach, the number of lanes open and closed, and the effective green time. Using such geometric and operating conditions that constitute work-zone environment, the proposed models successfully estimated SFR values with a miniscule margin of error.

• Journal of Broadcast Engineering
• /
• v.24 no.4
• /
• pp.580-591
• /
• 2019

This paper focuses on to improving object detection performance using the camera and LiDAR on autonomous vehicle platforms by fusing detected objects from individual sensors through a late fusion approach. In the case of object detection using camera sensor, YOLOv3 model was employed as a one-stage detection process. Furthermore, the distance estimation of the detected objects is based on the formulations of Perspective matrix. On the other hand, the object detection using LiDAR is based on K-means clustering method. The camera and LiDAR calibration was carried out by PnP-Ransac in order to calculate the rotation and translation matrix between two sensors. For Sensor fusion, intersection over union(IoU) on the image plane with respective to the distance and angle on world coordinate were estimated. Additionally, all the three attributes i.e; IoU, distance and angle were fused using logistic regression. The performance evaluation in the sensor fusion scenario has shown an effective 5% improvement in object detection performance compared to the usage of single sensor.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.41 no.6
• /
• pp.707-714
• /
• 2021

An important factor that causes delays at roundabouts is unbalanced traffic conditions. The majority of domestic studies have tendedto focus on the proportion of entering traffic rather than conflicting traffic in order to explain unbalanced traffic conditions. Also, there is scant research on the proper distance between a signalized intersection and a roundabout. In this study, therefore, unbalanced traffic conditions and optimal distance between two intersections were analyzed using SIDRA software based on the Gajwa-ro roundabout in Icheon, where unbalanced traffic conditions occur during afternoon peak hours, and its position is 295 m from a signalized intersection. It was found that a state of unbalanced traffic conditions in the form of conflicting traffic caused delays at Gajwa-ro roundabout, and the queuing length on each approach was removed when there was 850 m of distance between two intersections.

• Journal of Korean Neurosurgical Society
• /
• v.66 no.1
• /
• pp.53-62
• /
• 2023

Objective : Deep learning is a machine learning approach based on artificial neural network training, and object detection algorithm using deep learning is used as the most powerful tool in image analysis. We analyzed and evaluated the diagnostic performance of a deep learning algorithm to identify skull fractures in plain radiographic images and investigated its clinical applicability. Methods : A total of 2026 plain radiographic images of the skull (fracture, 991; normal, 1035) were obtained from 741 patients. The RetinaNet architecture was used as a deep learning model. Precision, recall, and average precision were measured to evaluate the deep learning algorithm's diagnostic performance. Results : In ResNet-152, the average precision for intersection over union (IOU) 0.1, 0.3, and 0.5, were 0.7240, 0.6698, and 0.3687, respectively. When the intersection over union (IOU) and confidence threshold were 0.1, the precision was 0.7292, and the recall was 0.7650. When the IOU threshold was 0.1, and the confidence threshold was 0.6, the true and false rates were 82.9% and 17.1%, respectively. There were significant differences in the true/false and false-positive/false-negative ratios between the anterior-posterior, towne, and both lateral views (p=0.032 and p=0.003). Objects detected in false positives had vascular grooves and suture lines. In false negatives, the detection performance of the diastatic fractures, fractures crossing the suture line, and fractures around the vascular grooves and orbit was poor. Conclusion : The object detection algorithm applied with deep learning is expected to be a valuable tool in diagnosing skull fractures.

• Korean Journal of Remote Sensing
• /
• v.39 no.5_1
• /
• pp.621-635
• /
• 2023

The assessment of structural condition is a crucial process for evaluating its usability and determining the diagnostic cycle. The currently employed manpower-based methods suffer from issues related to safety, efficiency, and objectivity. To address these concerns, research based on deep learning using images is being conducted. However, acquiring structural damage data is challenging, making it difficult to construct a substantial amount of training data, thus limiting the effectiveness of deep learning-based condition assessment. In this study, we propose a foundation model-based 2-step structural damage analysis to overcome the lack of training data in image-based structural condition assessments. We subdivided the elements of structural condition assessment into instantiation and quantification. In the quantification step, we applied a foundation model for image segmentation. Our method demonstrated a 10%-point increase in mean intersection over union compared to conventional image segmentation techniques, with a notable 40%-point improvement in the case of rebar exposure. We anticipate that our proposed approach will enhance performance in domains where acquiring training data is challenging.