본 연구에서는 콘크리트 이미지에서 균열의 크기와 위치를 검출하는 알고리즘을 개발하였다. 균열은 총 9단계로 자 동 검출되었으며, 기본 기능은 매트랩 프로그램의 기능이었다. 5단계와 8단계에서는 균열 검출 정확도를 높이기 위해 사용자 알고리즘을 추가하였으며, 균열 영상과 비균열 영상을 각각 1,000개씩 사용하였다. 균열 이미지에서는 균열이 100% 검출됐지만 품질 측면에서 나쁘지 않은 결과를 제외하더라도 91.8%의 결과가 매우 양호했다. 또한, 균열되지 않은 이미지의 정확도도 94.7%로 매우 양호했다. 이에 본 연구에서 제시한 균열검출 알고리즘은 콘크리트 우물 균열의 위치와 크기를 검출할 수 있을 것으로 기대된다.

Visual inspection methods have limitations, such as reflecting the subjective opinions of workers. Moreover, additional equipment is required when inspecting the high-rise buildings because the height is limited during the inspection. Various methods have been studied to detect concrete cracks due to the disadvantage of existing visual inspection. In this study, a crack detection technology was proposed, and the technology was objectively and accurately through AI. In this study, an efficient method was proposed that automatically detects concrete cracks by using a Convolutional Neural Network(CNN) with the Orthomosaic image, modeled with the help of UAV. The concrete cracks were predicted by three different CNN models: AlexNet, ResNet50, and ResNeXt. The models were verified by accuracy, recall, and F1 Score. The ResNeXt model had the high performance among the three models. Also, this study confirmed the reliability of the model designed by applying it to the experiment.

콘크리트 구조물 표면에 발생하는 균열은 사용자에게 심리적인 불안감을 제공하며, 장기간 열려있는 큰 폭의 균열은 구조 물의 사용성능 및 내구성에 영향을 준다. 국내에서는 건축물을 포함한 시설물의 노후화에 따른 안전관리를 위해 균열정도를 파악하는 조사가 인력에 의한 육안조사로 수행되고 있지만 인력의 고비용성과 객관성 미흡 등의 문제점이 대두되고 있다. 이를 해결하기 위해 영상분석을 통한 균열 추출 등 다양한 연구가 수행되고 있으나 균열인식 정확도 향상에 2차원 영상 분석만으로는 한계가 있다. 따라서, 본 연구에서는 기존 2차원 영상 분석의 한계를 극복하기 위하여 3차원 특성을 정확하게 파악할 수 있는 3차원 광삼각 스캐닝기법을 활용하여 콘크리트 구조물 표면의 균열정보를 획득하는 기법을 개발하였다. 본 하 드웨어의 개발과 더불어 균열 패턴분석을 위한 획득된 균열의 세분화와 균열의 특성분석 알고리즘을 개발하였으며, 이를 실제 콘크리트 빔의 균열 탐지 적용을 통해 검증하였다.

딥러닝 모델은 주어진 학습용 데이터에서 탐지하고자 하는 물체의 특징을 추출하기 때문에, 딥러닝 모델 학습을 위한 학습용 데이터 구축은 매우 중요하다. 본 연구에서는 균열을 탐지하는 딥러닝 모델의 성능을 향상시키기 위해, 실제 콘크리트 구조물이나 아스팔트 도로 표면에서 자주 발견될 수 있는 나뭇가지, 거미줄, 전선 등을 학습 데이터에 자동으로 포함시키고, negative 영역으로 분류하는 알고리즘을 개발하였다. 제안된 알고리즘을 사용하여 학습된 딥러닝 모델을 실제 도로 표면에 발생한 균열 탐지에 적용하여 실제 균열 탐지에 사용될 수 있음을 보였다.

최근 사회기반시설(SOC)의 증가와 노후화에 따라 기존의 인력중심의 육안검사를 기반으로 한 안전점검은 경제성과 안전성, 효율성 면에서 한계를 가지고 있다. 본 연구에서는 육안점검의 한계를 개선하기 위해 딥러닝 모델 기반 물체를 탐지하는 기술을 활용하여 터널 콘크리트 균열을 자동으로 탐지하는 기술을 개발하였으며, 이를 실제 터널 영상에 적용하여 그 성능을 검증하였다.

Last few years, many researches on deep learning-based crack detection model have been reported in order to develop an efficient structure inspection method. While developing crack detection deep learning model, many research results reported the importance of the training data. Since most of the research results only qualitatively discussed the importance of training data, this study examine the influence of the training data by experiment, especially in the case of negative samples such as construction joint, spider web and concrete blocks.

Recently, road maintenance is important for road performance and longevity in accordance with the increase of road infrastructure, so research on efficient damage detection has been actively carried out. In this study, we developed a technology to automatically detect the cracks on the highway road surface by using the road surface image of UAV and deep learning based object detection technology, which was applied to the actual highway image to verify its performance.

Pavement condition deteriorates due to various environmental issues. This can be seen on the pavement surface as a form of distress. A crack can be considered as a typical form of pavement distress in which it may reveal a critical condition of the road. Therefore, automatic and accurate detection of pavement crack and segmentation are crucial for pavement condition assessment and maintenance.

Acoustic Emission (AE) technique applied to detect the crack occurrence of the actual beam element. An optimum position for a limited number of AE transducer was considered to accurately detect the location of the cracks in the three-dimensional space. Six AE transducer was used to detect cracks in the L400mm×H200mm×T100mm region, and several position combinations were applied. Considerable six position combinations were selected, and the weak or incorrect position detection was investigated. The optimum position applied to the experiment for actual beam element and the detected crack position was compared with the visual inspection location. A reliable crack position detecting was confirmed for loadings.

In this study, we adapted very low resolution (240x320 = 76,800 pixels) images by which it is difficult to detect cracks. An automatic crack detection technique has been studied using digital image processing technology for low resolution images of sewage pipeline. Authors have developed two algorithms to detect cracks. The third step covers an algorithm developed to find optimal threshold value, and the sixth step deals with algorithm to determine cracks. As the result, in spite of very low-resolution images, the performance of crack detection turned out to be excellent.

Bridge inspection based unmanned aerial vehicle (UAV) has received considerable attention due to its several advantages such as reliability and safety as well as saving time and cost. An unmanned inspection equipment for bridge inspection is composed of UAV and imaging devices such as RGB cameras and infrared cameras. However, many challenging issues should be solved in order to apply this technology to the field. In this paper, an UAV based crack detection method is investigated. To detect the cracks, the image processing techniques with deep learning algorithm are used. To build the spatial information of aging bridge, 3D point cloud based background model is generated.

This paper proposes a deep learning-based crack evaluation technique using hybrid images. The use of the hybrid images combining vision and infrared images are able to improve crack detectability while minimizing false alarms. In particular, large-scale infrastructures can be inspected by an UAV-mounted hybrid image scanning (HIS)　system, and the corresponding huge amount of data is typically difficult to be analyzed by experts. To automate such making-decision process, deep convolutional neural network is used in this study. As the very first stage, a lab-scale HIS system is developed using a scanning zig and experimentally validated using a concrete specimen with various-size cracks. The test results reveal that macro- and micro-cracks are successfully and automatically detected with minimizing false-alarms.