• Smart Structures and Systems
• /
• v.31 no.4
• /
• pp.325-334
• /
• 2023

Computer vision-based damage detection enables non-contact, efficient and low-cost bridge health monitoring, which reduces the need for labor-intensive manual inspection or that for a large number of on-site sensing instruments. By leveraging recent semantic segmentation approaches, we can detect regions of critical structural components and identify damages at pixel level on images. However, existing methods perform poorly when detecting small and thin damages (e.g., cracks); the problem is exacerbated by imbalanced samples. To this end, we incorporate domain knowledge to introduce a hierarchical semantic segmentation framework that imposes a hierarchical semantic relationship between component categories and damage types. For instance, certain types of concrete cracks are only present on bridge columns, and therefore the noncolumn region may be masked out when detecting such damages. In this way, the damage detection model focuses on extracting features from relevant structural components and avoid those from irrelevant regions. We also utilize multi-scale augmentation to preserve contextual information of each image, without losing the ability to handle small and/or thin damages. In addition, our framework employs an importance sampling, where images with rare components are sampled more often, to address sample imbalance. We evaluated our framework on a public synthetic dataset that consists of 2,000 railway bridges. Our framework achieves a 0.836 mean intersection over union (IoU) for structural component segmentation and a 0.483 mean IoU for damage segmentation. Our results have in total 5% and 18% improvements for the structural component segmentation and damage segmentation tasks, respectively, compared to the best-performing baseline model.

• International conference on construction engineering and project management
• /
• 2013.01a
• /
• pp.279-286
• /
• 2013

This paper introduces a new method for identification of building energy performance problems. The presented method is based on automated analysis and visualization of deviations between actual and expected energy performance of the building using EPAR (Energy Performance Augmented Reality) models. For generating EPAR models, during building inspections, energy auditors collect a large number of digital and thermal imagery using a consumer-level single thermal camera that has a built-in digital lens. Based on a pipeline of image-based 3D reconstruction algorithms built on GPU and multi-core CPU architecture, 3D geometrical and thermal point cloud models of the building under inspection are automatically generated and integrated. Then, the resulting actual 3D spatio-thermal model and the expected energy performance model simulated using computational fluid dynamics (CFD) analysis are superimposed within an augmented reality environment. Based on the resulting EPAR models which jointly visualize the actual and expected energy performance of the building under inspection, two new algorithms are introduced for quick and reliable identification of potential performance problems: 1) 3D thermal mesh modeling using k-d trees and nearest neighbor searching to automate calculation of temperature deviations; and 2) automated visualization of performance deviations using a metaphor based on traffic light colors. The proposed EPAR v2.0 modeling method is validated on several interior locations of a residential building and an instructional facility. Our empirical observations show that the automated energy performance analysis using EPAR models enables performance deviations to be rapidly and accurately identified. The visualization of performance deviations in 3D enables auditors to easily identify potential building performance problems. Rather than manually analyzing thermal imagery, auditors can focus on other important tasks such as evaluating possible remedial alternatives.

• Korean Journal of Remote Sensing
• /
• v.40 no.1
• /
• pp.93-101
• /
• 2024

Accurate field crop classification is essential for various agricultural applications, yet existing methods face challenges due to diverse crop types and complex field conditions. This study aimed to address these issues by combining support vector machine (SVM) models with multi-seasonal unmanned aerial vehicle (UAV) images, texture information extracted from Gray Level Co-occurrence Matrix (GLCM), and RGB spectral data. Twelve high-resolution UAV image captures spanned March-October 2021, while field surveys on three dates provided ground truth data. We focused on data from August (-A), September (-S), and October (-O) images and trained four support vector classifier (SVC) models (SVC-A, SVC-S, SVC-O, SVC-AS) using visual bands and eight GLCM features. Farm maps provided by the Ministry of Agriculture, Food and Rural Affairs proved efficient for open-field crop identification and served as a reference for accuracy comparison. Our analysis showcased the significant impact of hyperparameter tuning (C and gamma) on SVM model performance, requiring careful optimization for each scenario. Importantly, we identified models exhibiting distinct high-accuracy zones, with SVC-O trained on October data achieving the highest overall and individual crop classification accuracy. This success likely stems from its ability to capture distinct texture information from mature crops.Incorporating GLCM features proved highly effective for all models,significantly boosting classification accuracy.Among these features, homogeneity, entropy, and correlation consistently demonstrated the most impactful contribution. However, balancing accuracy with computational efficiency and feature selection remains crucial for practical application. Performance analysis revealed that SVC-O achieved exceptional results in overall and individual crop classification, while soybeans and rice were consistently classified well by all models. Challenges were encountered with cabbage due to its early growth stage and low field cover density. The study demonstrates the potential of utilizing farm maps and GLCM features in conjunction with SVM models for accurate field crop classification. Careful parameter tuning and model selection based on specific scenarios are key for optimizing performance in real-world applications.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.3
• /
• pp.87-94
• /
• 2024

We investigated the feasibility of utilizing Large Language Models (LLMs) to perform text-based games without training on game data in advance. We adopted ChatGPT-3.5 and its state-of-the-art, ChatGPT-4, as the systems that implemented LLM. In addition, we added the persistent memory feature proposed in this paper to ChatGPT-4 to create three game player agents. We used Zork, one of the most famous text-based games, to see if the agents could navigate through complex locations, gather information, and solve puzzles. The results showed that the agent with persistent memory had the widest range of exploration and the best score among the three agents. However, all three agents were limited in solving puzzles, indicating that LLM is vulnerable to problems that require multi-level reasoning. Nevertheless, the proposed agent was still able to visit 37.3% of the total locations and collect all the items in the locations it visited, demonstrating the potential of LLM.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.15 no.2
• /
• pp.91-98
• /
• 2009

The long-term water-quality change of Asan Bay by the influx of polluted disposal water was predicted through a simulation with an Eco-hydrodynamic model. Eco-hydrodynamic model is composed of a multi-level hydrodynamic model to simulate the water flow and an ecosystem model to simulate water quality. The water quality simulation revealed that the COD(Chemical Oxygen Demand), dissolved inorganic nitrogen(DIN) and dissolved inorganic phosphorus(DIP) are increased at 5 stations for the subsequent 6 months after the influx of the effluent. COD, DIN and DIP showed gradual decreases in concentration during the period of one to two years after the increase of last 6 months and reached steady state for next three to ten years. Concentration levels of COD, DIN, and DIP showed the increase by the ranges of $11{\sim}67%$, $10{\sim}67%$, and $0.5{\sim}7%$, respectively, which represents that the COD and DIN are the most prevalent pollutants among substances in the effluent through the sewage treatment plant. The current water quality of Asan Bay based on the observed COD, TN and TP concentrations ranks into the class II of the Korean standards for marine water quality but the water quality would deteriorate into class III in case that the disposal water by the sewage plant is discharged into the Bay.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.36 no.3
• /
• pp.277-284
• /
• 2012

In this study, a heat-transfer performance analysis is carried out for a multi-channel volumetric air receiver for a solar power tower. On the basis of a series of reviews regarding the relevant literature, a calculation process is proposed for the prediction of the wall- and air- temperature distributions of a single channel at given geometric and input conditions. Furthermore, a unique mathematical model of the receiver effectiveness is presented through analysis of the temperature profile. The receiver is made of silicon carbide. A total of 225 square straight channels per module are molded to induce the air flow, and each channel has the dimensions of $2mm(W){\times}2mm(H){\times}0.2mm(t){\times}320mm(L)$. The heat-transfer rate, temperature distribution and effectiveness are presented according to the variation of the channel and module number under uniform irradiation and mass flow rate. The available air outlet temperature applied to the solar power tower should be over $700^{\circ}C$. This numerical model was actually used in the design of a 200 kW-level commercial solar air receiver, and the required number of modules satisfying the thermal performance could be obtained for the specified geometric and input conditions.

• Journal of Korean Society of Transportation
• /
• v.36 no.3
• /
• pp.169-183
• /
• 2018

Aggressive driving leads to a greater crash potential because it threatens surrounding vehicles. This study conducted traffic simulation experiments using driving behavior data obtained from multi-agent driving simulations. VISSIM traffic simulator and surrogate safety assessment model (SSAM) were used to identify the impacts of aggressive driving on traffic stream in terms of safety and operational efficiency. Market penetration rates (MPR) of aggressive driving vehicle, coupled with various traffic conditions, were taken into consideration in analyzing the impacts. As expected, it was identified that aggressive driving vehicles tended to deteriorate the traffic safety performance. From the perspective of operational efficiency, interesting results were observable. Under level of service (LOS) A, B, and C, it was observed that the average travel speed increased with greater MPRs. Conversely, the average travel speed decreased with under LOS D and E conditions. The outcome of this study would be effectively used for developing safety-related policies for reducing aggressive driving behavior.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.8
• /
• pp.96-105
• /
• 1998

The aim of this paper is to propose the method of WDHMM(Weighted DHMM), using the MLP-VQ for the improvement of speaker-independent connect-digit recognition system. MLP neural-network output distribution shows a probability distribution that presents the degree of similarity between each pattern by the non-linear mapping among the input patterns and learning patterns. MLP-VQ is proposed in this paper. It generates codewords by using the output node index which can reach the highest level within MLP neural-network output distribution. Different from the old VQ, the true characteristics of this new MLP-VQ lie in that the degree of similarity between present input patterns and each learned class pattern could be reflected for the recognition model. WDHMM is also proposed. It can use the MLP neural-network output distribution as the way of weighing the symbol generation probability of DHMMs. This newly-suggested method could shorten the time of HMM parameter estimation and recognition. The reason is that it is not necessary to regard symbol generation probability as multi-dimensional normal distribution, as opposed to the old SCHMM. This could also improve the recognition ability by 14.7% higher than DHMM, owing to the increase of small caculation amount. Because it can reflect phone class relations to the recognition model. The result of my research shows that speaker-independent connected-digit recognition, using MLP-VQ and WDHMM, is 84.22%.

• International Area Studies Review
• /
• v.15 no.3
• /
• pp.95-126
• /
• 2011

The objective of this study is to quantify the potential economic effects of Non-Agricultural Market Access (NAMA) negotiations of the WTO on automotive industry of the world using a multi-region, multi-sector Computable General Equilibrium (CGE) model with 21 countries/regions and 22 sectors. According to the December 2008 NAMA modalities text, issued by the chair of the negotiation on NAMA, three different scenarios of tariff liberalization of NAMA are conducted on the basis of the Swiss formula with a coefficient of 8 for developed members and 20 for developing (scenario 1), with a coefficient of 8 for developed members and 22 for developing (scenario 2) and with a coefficient of 8 for developed members and 25 for developing (scenario 3). Simulation results show potential economic effects at the macroeconomic and microeconomic level of 21 countries concerned. In particular, Korea is to be one of the winners of tariff liberalization of NAMA in the WTO and Korean automotive industry is to benefit from it to a large extent in terms of its output, domestic sales, exports and trade balance, which implies that Korea needs to actively engage in NAMA negotiations of the WTO.

• Korean Journal of Social Welfare Studies
• /
• v.42 no.3
• /
• pp.183-206
• /
• 2011

This study examined the longitudinal relationship between academic achievement and school satisfaction using a data of the Seoul Panel Study of Children(SPSC). Fully autoregressive cross-lagged analysis and multi-group comparison were performed to measure the longitudinal relationship between two constructs as well as differences between poverty and non-poverty groups. The results showed that both academic achievement and school satisfaction were stable over time in non-poverty group. Academic achievement at the 4th grade significantly affected the school satisfaction at the 6th grade and it subsequently affected on the academic achievement at the 8th grade in non-poverty group. In contrast, academic achievement was not consistent over time in poverty group. Only the school satisfaction at the 6th grade affected the academic achievement at the 8th grade. The findings of this study have various practical implication for school interventions. It is more important to keep supporting the children to maintain the level of academic achievement in non-poverty group. While, in poverty group, it is essential to make school satisfaction and academic motivation increase with school attachment programs.