• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.53-56
• /
• 2021

국내 산업화가 들어선 후 산업화 당시 지었던 낙후된 건물의 증가에 따라 구조물의 손상 조사 및 검사 방법의 수요가 늘어나고 있다. 일반적으로 구조물의 손상은 전문 검사원이 현장에서 직접 측량도구와 시각적인 방식으로 검사한다. 그러나 전문 검사원들이 직접 조사하는 수고에 비해 균열을 검사하는 방식 자체가 단순하고, 일반 사람이 검사하기에는 객관성이 떨어지는 한계가 있어 균열을 자동적으로 검출함으로써 객관성과 편의성을 보장할 기술이 필요하다. 본 연구에서는 이미지 기반으로 다양한 환경에서의 외벽 균열을 검출할 수 있는 딥러닝 모델 개발을 소개한다. 균열 검출을 위해 다양한 외벽 균열 관련 데이터셋을 확보 및 구축하고 각 데이터셋의 검출 정보를 보완할 반자동(semi-auto) 라벨링 작업을 수행하였다. 두 번째로 기존 높은 검출 성능을 보였던 모델들을 선정 및 비교하여 YOLO v5 모델을 최종적으로 선정하였고, 도메인이 각각 다른 데이터셋에 대한 교차 학습을 통해 각 데이터셋의 mAP의 편차가 31%에서 11%로 좁히는 작업을 수행하였다. 이를 통해 실제 상황에서의 균열 영상에서 균열을 검출할 수 있는 측량 시스템을 개발함으로써 실질적인 검사의 도구로 활용될 수 있길 기대한다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.49-52
• /
• 2022

건물 외벽에 발생하는 균열은 시설물 구조 안전에 영향을 미치며 그 크기에 따라 위험도가 달라진다. 이에 따라 전문검사관의 현장 점검을 통해 발생 균열 두께를 정밀하게 측정할 필요가 있고 최근에는 이러한 현장 안전점검에 인공지능을 도입하려는 추세다. 그러나 기존의 균열 데이터셋은 주로 콘크리트에만 한정되어 다양한 외벽에 강인한 모델을 구축하기 어렵고 균열 두께를 측정하기 위해 정확한 마스크(Mask) 정보가 필요하나 이를 만족하는 데이터셋이 부재하다. 본 논문에서는 다양한 외벽에 강인한 균열 구획화 모델을 목적으로 2,744장의 이미지를 촬영하고 매직 완드 기법으로 라벨링을 진행해 데이터셋을 구축 후, 이를 바탕으로 딥러닝 기반 균열 구획화 모델을 개발했다. UNet-ResNet50을 최종모델로 선정 및 개발 결과, 테스트 데이터셋에 대해 81.22%의 class IoU 성능을 보였다. 본 연구의 기술을 바탕으로 균열 두께를 측정하여 건축물 안전점검에 활용될 수 있기를 기대한다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.5
• /
• pp.167-178
• /
• 2020

The purpose of this paper is to establish a defect classification system for defects before inspection and to derive the pattern and characteristics of defects before inspection by examining about 3,110 defect items for 133 apartment buildings. The study analysis revealed a relatively high rate of defects before inspection that occurred in finishing work. Second, defects occurred such as cracking of external wall, which is a very important defect. However, defects before inspection were relatively rare on the external wall. Finally, defects before inspection occurred during waterproofing in the common area or garage. It is necessary to establish a reasonable basis or countermeasure to resolve differences between stakeholders as various issues may arise in the course of a dispute, as a result of identifying the details of defects within the top 20 of the defectives.

• Journal of the Korean Society for Precision Engineering
• /
• v.30 no.4
• /
• pp.359-367
• /
• 2013

With recent growth of the economy and development of construction technology, the increase of high-rise buildings is appearing rapidly in urban areas. For this reason, the interest in building maintenance has also been increasing. However, it has many safety problems because it is difficult for the workers to access the exterior wall of building. Therefore, the maintenance system of high-rise building stands out as being important issues to be developed, so that a variety of robot systems have been developed to accomplish the building-wall maintenance works. In this paper, the maintenance robots are classified in painting, inspecting, cleaning systems according to the maintenance works. Then, their locomotion and adhesion mechanisms are analyzed including their applicability to the real maintenance works. This study can be used to develop maintenance robotic system that is more efficient and stable than existing ones.

• Proceedings of the Korean Society of Disaster Information Conference
• /
• 2023.11a
• /
• pp.219-220
• /
• 2023

본 논문에서는 건축물의 고층화, 대형화로 외벽 청소, 도장, 견출, 검사등 작업시 손쉽고, 빠르며 경제적으로 작업을 위한 방법으로 작업의자형 달비계를 사용 하고 있으나 최근 3년간 유지·보수공사 작업시 근로자의 중대재해가 36명으로 신축·증축·재건축공사 2명으로 유지·보수공사에서 많은 중대재해가 발생되고 있는 실정으로 50억이상 신축 작업에서는 산업안전보건법에 의하여 관리감독자 및 안전관리자가 선임 되어 현장의 전반적인 안전관리 활동을 하고 있으나 50억 이하 소규모 작업장에서는 안전관리 할동이 전무한 실정이고 저가수주, 적정 공사기간 불가 및 산업안전보건관리비 미계상 등 근로자들에 대한 안전대책이 없고 외부 안전점검기관의 사각지대로 근로자의 안전 확보에 어려움이 있은 실정으로 사업주 및 근로자에 대한 안전교육 및 관리감독이 철저히 이루어져야 할 것이다

• Proceedings of the KIEE Conference
• /
• 2015.07a
• /
• pp.868-869
• /
• 2015

지하 매설된 가스배관을 정기적으로 검사하기 위해서 가스 공급 및 용역업체에서는 주로 비피과검사 탐상장비인 MFL(Magnetic Flux Leakage) PIG(Pipeline Inspection Gauge)를 사용한다. 기존의 MFL PIG는 배관 내 유체(가스,오일 등)의 전후차단 압력의 흐름을 이용해 별도의 구동장치 없이 피그를 진행시켜 배관의 결함 유무를 평가하는 시스템이다. 하지만 10기압 이하의 낮은 운영압력과 T 분기관과 같이 급격한 곡관부가 존재하는 직경 16인치 이하의 도시가스 배관에는 기존의 시스템을 적용하기 어렵다. 이처럼 기존 MFL PIG의 적용이 불가한 도시가스 배관(직경 16인치 이하)을 활주하기 위해서는 우선 비파괴검사 시스템을 견인할 수 있는 추진 로봇이 필요하고 추진로봇에 적합한 자기누설 비파괴검사 시스템의 개발이 필요하다. 또한 비파괴검사 장비의 센서 시스템은 결함신호를 탐지하여 결함의 발생유무 및 결함의 형상을 판별하는 성능도 중요하다. 본 논문에서는 16인치 도시가스 배관의 검사를 위한 자기누설 비파괴검사 시스템의 기초설계와 대상 시스템의 자기적 특성을 분석한다. 또한 배관 외벽의 결함 발생 유무에 따른 자기누설 신호의 크기 및 분포변화를 3차원 유한요소법을 이용해 해석하여, 결함 검출 신호의 특성을 분석하는데 초점을 둔다.

• Clinical and Experimental Pediatrics
• /
• v.50 no.5
• /
• pp.476-483
• /
• 2007

Purpose : The objective of this study was to assess ventricular function by tissue Doppler imaging (TDI) in children with congenital heart disease (CHD) who have been undergoing open heart surgery (OHS) using cardiopulmonary bypass. We tried to compare the parameters of tissue Doppler imaging before and after OHS in patients with congenital heart disease. Methods : This study was conducted on 32 patients with CHD after OHS from January 2005 to December 2005 at Kyungpook National University hospital. Patients who underwent 2-D echocardiography before and after their OHS. All patients were divided into three groups, left ventricular volume overloading group (group 1), and right ventricular volume overloading group (group 2), and right ventricular pressure overloading group (group 3). The TDIs were examined before and 1 to 3 months after OHS. Peak early diastolic (E), and peak late diastolic (A) velocity of transmitral flow were measured by pulsed wave Doppler examination. Peak systolic (Sm), peak early diastolic (Em), and peak late diastolic (Am) velocity in apical 4-chamber and 2-chamber views were measured by TDI. The author calculated E/Em ratio. Results : The patients were 14 boys and 18 girls and the average age of patients was 2 years and 3 months. The congenital heart diseases which have to get OHS were ventricular septal defect (13 cases), atrial septal defect (7), atrioventricular septal defect (3), isolated pulmonary stenosis (2) and tetralogy of Fallot (7). There were significant decrease of Sm, Em, Am measured on tricuspid annulus and E/Em measured on mitral annulus in apical 4 chamber view (P<0.05). Conclusion : This study showed significant decrease of Sm, Em, Am measured on tricuspid annulus and E/Em measured on mitral annulus in apical 4 chamber view after OHS. These changes might be due to the effects of cardiopulmonary bypass in OHS and/or hemodynamic changes after correction of congenital heart disease. To clarify these changes, further study on more patients is needed.

• KEPCO Journal on Electric Power and Energy
• /
• v.5 no.1
• /
• pp.33-38
• /
• 2019

Thermal power generation accounts for the highest percentage of domestic power generation, among which coal-fired boiler generation accounts for the highest percentage. Coal boilers generate harmful substances and fine dust during coal combustion and have a serious effect on air pollution. So, fluidized-bed boilers have been introduced as eco-friendly coal boilers. It uses a fluid medium which affect the combustion temperature of coal. Because of it fluidized-bed boilers emit less pollutants than original one. Water-wall tubes play an important role in this fluidized bed boiler. Due to the fluid medium, the wall damage is more severe than the existing boiler. However, there is no quantitative maintenance technique in Korea yet. Remote field eddy current testing is a non-destructive evaluation technique that is often used for inspection of inner and outer wall of tube. it can inspect with non-contact and high speed. However, it is an inspection that proceeds from inside the pipe, and the water-wall tube is not able to enter the interior. In this study, we designed and simulated an external remote field eddy current sensor suitable for water-wall tube of a fluidized - bed boiler using simulations. By obtaining a signal similar to the existing remote field eddy current test, the criteria for the external remote field eddy current sensor design can be presented.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.22 no.2
• /
• pp.187-192
• /
• 2002

The open-connect type eddy current transducer for the flaw detection in continuously connected pipelines was developed. This eddy current transducer is for the on-line inspection of the tubes in industries, to which commercial encircling probes are not applicable. The excitation coil that consists of a ribbon type cable and a flat connector can be opened and closed on purpose. The sensing coils of this transducer are circumferentially arrayed near the outside of the tube wall but axially displaced from the exciter by about one and half tube diameter. In application to steel tubes, and the performance of this transducer was evaluated as a little behind those of magnetic saturation type in signal to noise ratio and flaw size decision, but usable to detect or to locate large size flaws in steel tubes. Surface cracks deeper than 19% of the tube thickness could be detected with good signal to noise ratio.

• Journal of the Korea Institute of Building Construction
• /
• v.23 no.2
• /
• pp.197-207
• /
• 2023

Concrete is a widely used material due to its excellent compressive strength and durability. However, depending on the surrounding environment and the characteristics of the materials used in the construction, various defects may occur, such as cracks on the surface and subsidence of the structure. The detects on the surface of the concrete structure occur after completion or over time. Neglecting these cracks may lead to severe structural damage, necessitating regular safety inspections. Traditional visual inspections of concrete walls are labor-intensive and expensive. This research presents a deep learning-based semantic segmentation model designed to detect cracks in concrete walls. The model addresses surface defects that arise from aging, and an image augmentation technique is employed to enhance feature extraction and generalization performance. A dataset for semantic segmentation was created by combining publicly available and self-generated datasets, and notable semantic segmentation models were evaluated and tested. The model, specifically trained for concrete wall fracture detection, achieved an extraction performance of 81.4%. Moreover, a 3% performance improvement was observed when applying the developed augmentation technique.