• 한국컴퓨터정보학회논문지
• /
• 제21권6호
• /
• pp.71-76
• /
• 2016

In this paper, we propose the mobile visual inspection information system using DGPS and portable range finder for levee safety inspection. Instead of existing visual inspection management method that is stored hand-written data, this system is designed to manage directly the visual inspection information using mobile devices in the field of levee. And through extracting accurate DGPS coordinates information about damage location of levee, this system is developed to ensure efficiency for the main task arising from the levee site such as inspection, maintenance and reinforcement. Furthermore, when damage has occurred at the point that inspector is not able to approach, this system can record the damage site data correctly, by converting data such as position, orientation and height of the damage point into the World Geodetic System coordinates. The position, orientation and height data was extracted automatically through the DGPS and portable range finder. And by applying the augmented reality method, this system was implemented for inspector to revisit the point of damage easily in order to perform the management, maintenance and reinforcement of the levee later.

• Smart Structures and Systems
• /
• 제4권5호
• /
• pp.605-628
• /
• 2008

In bridge structures, damage may induce an additional deflection which may naturally contain essential information about the damage. However, inverse mapping from the damage-induced deflection to the actual damage location and severity is generally complex, particularly for statically indeterminate systems. In this paper, a new load concept, called the positive-bending-inspection-load (PBIL) is proposed to construct a simple inverse mapping from the damage-induced deflection to the actual damage location. A PBIL for an inspection region is defined as a load or a system of loads which guarantees the bending moment to be positive in the inspection region. From the theoretical investigations, it was proven that the damage-induced chord-wise deflection (DI-CD) has the maximum value with the abrupt change in its slope at the damage location under a PBIL. Hence, a novel damage localization method is proposed based on the DI-CD under a PBIL. The procedure may be summarized as: (1) identification of the modal flexibility matrices from acceleration measurements, (2) design for a PBIL for an inspection region of interest in a structure, (3) calculation of the chord-wise deflections for the PBIL using the modal flexibility matrices, and (4) damage localization by finding the location with the maximum DI-CD with the abrupt change in its slope within the inspection region. Procedures from (2)-(4) can be repeated for several inspection regions to cover the whole structure complementarily. Numerical verification studies were carried out on a simply supported beam and a three-span continuous beam model. Experimental verification study was also carried out on a two-span continuous beam structure with a steel box-girder. It was found that the proposed method can identify the damage existence and damage location for small damage cases with narrow cuts at the bottom flange.

• International Journal of Quality Innovation
• /
• 제7권2호
• /
• pp.97-124
• /
• 2006

The purpose of this research is to create an expert risk-based piping system inspection model. The proposed system includes a risk-based piping inspection system and a piping inspection guideline system. The research procedure consists of three parts: the risk-based inspection model, the risk-based piping inspection model, and the piping inspection guideline system model. In this research procedure, a field plant visit is conducted to collect the related domestic information (Taiwan) and foreign standards and regulations for creating a strategic risk-based piping inspection and analysis system in accordance with the piping damage characteristics in the petrochemical industry. In accordance with various piping damage models and damage positions, petrochemical plants provide the optimal piping inspection planning tool for efficient piping risk prediction for enhancing plant operation safety.

• 대한토목학회논문집
• /
• 제38권4호
• /
• pp.621-625
• /
• 2018

본 연구는 딥러닝을 활용하여 교량 점검보고서에서 손상 및 손상 인자를 자동으로 식별하는 방법을 제안한다. 교량 점검보고서에는 점검 결과 발견된 손상 및 원인 분석 결과가 기록되어 있다. 그러나 점검보고서의 양이 방대하여 인력으로 보고서로부터 정보를 수집하는 데 한계가 있다. 따라서 본 연구에서는 딥러닝 기반 개체명 인식 방법을 활용하여 교량 점검보고서 텍스트로부터 손상 및 손상 인자에 해당하는 단어들을 식별할 수 있는 모델을 제안한다. 모델 구현의 주요 방법론으로는 개체명 인식(Named Entity Recognition), 워드 임베딩(Word Embedding), 딥러닝의 일종인 순환신경망(Recurrent Neural Network)을 활용하였다. 실험 결과 제안된 모델은 1)훈련 데이터에 포함된 손상 및 손상 인자 단어들을 잘 식별할 수 있고, 2)단어 주변 맥락에 따라 특정 단어가 손상에 해당하는지 손상 인자에 해당하는지 잘 판별할 수 있을 뿐만 아니라, 3)훈련 데이터에 포함되지 않은 새로운 종류의 손상 단어도 잘 인식할 수 있는 것으로 확인되었다.

• Smart Structures and Systems
• /
• 제31권6호
• /
• pp.573-583
• /
• 2023

Although unmanned aerial vehicles have been used to overcome the limited accessibility of human-based visual inspection, unresolved issues still remain. Onsite inspectors face difficulty finding previously detected damage locations and tracking their status onsite. For example, an inspector still marks the damage location on a target structure with chalk or drawings while comparing the current status of existing damages to their previous status, as documented onsite. In this study, an augmented-reality-based structural inspection system with onsite damage information marking was developed to enhance the convenience of inspectors. The developed system detects structural damage, creates a holographic marker with damage information on the actual physical damage, and displays the marker onsite via an augmented reality headset. Because inspectors can view a marker with damage information in real time on the display, they can easily identify where the previous damage has occurred and whether the size of the damage is increasing. The performance of the developed system was validated through a field test, demonstrating that the system can enhance convenience by accelerating the inspector's essential tasks such as detecting damages, measuring their size, manually recording their information, and locating previous damages.

• 대한토목학회논문집
• /
• 제37권1호
• /
• pp.19-28
• /
• 2017

이 논문에서는, 1996년부터 2010년까지 915개 고속도로 교량에 대해 실시된 정밀안전진단 보고서를 수집하여 이들 교량의 상태를 분석하였다. 분석시 손상을, 결함, 물리력 및 열화에 의한 손상으로 구분하여 손상이 많이 발생하는 시기를 조사하고, 유손상율의 개념을 도입하여 손상의 특징을 조사하였으며, 고속도로 교량의 10대 손상, 다설한랭지역 교량과 일반지역 교량의 열화특성을 비교하여 합리적인 점검활동 개선 방향을 제시하였다. 연구결과, 시공시 실시하는 점검 또는 초기점검 제도 개선이 필요하고, 다설한랭지역의 경우 열화진행 속도가 빠르므로 주변 환경의 특성을 고려한 점검 활동 개선이 필요한 것으로 판단되었다. 이 연구 결과는 향후 고속도로 교량의 점검 활동 개선을 위해 널리 활용될 수 있을 것이다.

• 한국정밀공학회지
• /
• 제22권1호
• /
• pp.82-88
• /
• 2005

The damage inspection of container surface is performed by the expert inspectors at the container terminal gate of harbor. In this paper, we substitute the expert's capability with the damage inspection system using the artificial intelligent control algorithm and vision system, so we can improve the work environment and effectively decrease the inspection time and cost. Firstly, using six CCD cameras attached to the terminal gate, whole container is partially captured according to eleven sensors aligned with the entering direction of container. Captured partial images are inspected by the fuzzy system which the expert's technology is embedded. Next, we compose partial images to be a complete container image through the correlation coefficient method. Complete container image is saved to solve future troublesome problems. In this paper, the effectiveness of the proposed system was verified through the field test.

• 한국추진공학회:학술대회논문집
• /
• 한국추진공학회 2006년도 제26회 춘계학술대회논문집
• /
• pp.239-245
• /
• 2006

초음파를 이용한 추진기관의 비파괴검사는 X-ray 검사에 비하여 경제성이 우수하고, X-ray 검사시 취약한 dis-bond, damage 등의 결함 검출이 우수한 편이다. 그리고 전용시설의 필요없이 현장에서 실시간으로 검사가 가능하며 방사선 작업에 비하여 안전한 방법이다. 본 논문에서는 고체 추진제에 대한 초음파 특성을 분석하고, 추진제/라이너 미접착에 대한 내측과 외측 검사 방법 및 추진제 damage에 의한 미세 크랙검출에 대하여 실험 및 분석하였다. 실험 결과, 추진제에서의 초음파 감쇠는 $6\sim8db/cm$로 비교적 큰 감쇠를 보였으며 추진제/라이너 미접착에 대한 내 외측 검사에 있어서도 제한된 조건이지만 검출 가능성을 보였다. 그리고 damage에 의한 추진제 미세 크랙도 초음파의 감쇠특성을 이용하여 검출 가능함을 보였다.

• Smart Structures and Systems
• /
• 제25권6호
• /
• pp.733-749
• /
• 2020

This study presents a computer vision-based approach for representing time evolution of structural damages leveraging a database of inspection images. Spatially incoherent but temporally sorted archival images captured by robotic cameras are exploited to represent the damage evolution over a long period of time. An access to a sequence of time-stamped inspection data recording the damage growth dynamics is premised to this end. Identification of a structural defect in the most recent inspection data set triggers an exhaustive search into the images collected during the previous inspections looking for correspondences based on spatial proximity. This is followed by a view synthesis from multiple candidate images resulting in a single reconstruction for each inspection round. Cracks on concrete surface are used as a case study to demonstrate the feasibility of this approach. Once the chronology is established, the damage severity is quantified at various levels of time scale documenting its progression through time. The proposed scheme enables the prediction of damage severity at a future point in time providing a scope for preemptive measures against imminent structural failure. On the whole, it is believed that the present study will immensely benefit the structural inspectors by introducing the time dimension into the autonomous condition assessment pipeline.

• Smart Structures and Systems
• /
• 제31권4호
• /
• pp.335-349
• /
• 2023

Bridges constantly undergo deterioration and damage, the most common ones being concrete damage and exposed rebar. Periodic inspection of bridges to identify damages can aid in their quick remediation. Likewise, identifying components can provide context for damage assessment and help gauge a bridge's state of interaction with its surroundings. Current inspection techniques rely on manual site visits, which can be time-consuming and costly. More recently, robotic inspection assisted by autonomous data analytics based on Computer Vision (CV) and Artificial Intelligence (AI) has been viewed as a suitable alternative to manual inspection because of its efficiency and accuracy. To aid research in this avenue, this study performs a comparative assessment of different architectures, loss functions, and ensembling strategies for the autonomous segmentation of bridge components and damages. The experiments lead to several interesting discoveries. Nested Reg-UNet architecture is found to outperform five other state-of-the-art architectures in both damage and component segmentation tasks. The architecture is built by combining a Nested UNet style dense configuration with a pretrained RegNet encoder. In terms of the mean Intersection over Union (mIoU) metric, the Nested Reg-UNet architecture provides an improvement of 2.86% on the damage segmentation task and 1.66% on the component segmentation task compared to the state-of-the-art UNet architecture. Furthermore, it is demonstrated that incorporating the Lovasz-Softmax loss function to counter class imbalance can boost performance by 3.44% in the component segmentation task over the most employed alternative, weighted Cross Entropy (wCE). Finally, weighted softmax ensembling is found to be quite effective when used synchronously with the Nested Reg-UNet architecture by providing mIoU improvement of 0.74% in the component segmentation task and 1.14% in the damage segmentation task over a single-architecture baseline. Overall, the best mIoU of 92.50% for the component segmentation task and 84.19% for the damage segmentation task validate the feasibility of these techniques for autonomous bridge component and damage segmentation using RGB images.