Recognition of Damaged Building Using Deep Learning Based on Aerial and Local Photos Taken After the 1995 Kobe Earthquake

Ishii, Yuu; Matsuoka, M.; Maki, Norio; Horie, Kei; Tanaka, Satoshi

doi:10.3130/aijs.83.1391

Cited by 10 publications

(7 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…12 CNN-based image classification has also been proposed to evaluate the damage degrees of building structures from field survey photos. 5,13 On the other hand, the evaluation of mechanical property degradation for seismically damaged RC components has gained attention, [14][15][16] where observable or inferable parameters such as maximum residual crack width and ductility coefficient were used as damage indicators. However, among studies of vision-based damage detection and evaluation of mechanical property degradation for RC components, there is a lack of research on the connection of these two aspects.…”

Section: F I G U R E 1 Proposed Evaluation Approach For the Mechanica...mentioning

confidence: 99%

Deep learning‐based evaluation for mechanical property degradation of seismically damaged RC columns

Miao

et al. 2022

Earthq Engng Struct Dyn

View full text Add to dashboard Cite

The evaluation of mechanical property degradation (i.e., stiffness and strength degradation) for seismically damaged reinforced concrete (RC) components is a critical step in the post-earthquake assessment of the residual seismic capacity of buildings. In this study, a novel approach based on deep learning (DL) was proposed to evaluate the stiffness and strength degradation of RC columns according to visible seismic damage. A database was constructed by linking the test photos of RC column specimens with the loading points on the hysteretic curves, from which the stiffness and strength reduction factors (𝜆 K and 𝜆 Q , respectively) were analyzed. Two novel convolutional network (CNN) modules were designed to enable feature extraction and integration of seismic damage with a reduced number of parameters, and multitask learning was introduced to enable adaptive feature fusion for stiffness and strength degradation individually.A deep convolutional network (DCNN) was therefore proposed to model the correlation between visible seismic damage and mechanical property degradation of flexural-dominated RC columns, which can integrate visual characteristics and spatial topologies of visible damage to estimate 𝜆 K and 𝜆 Q . The application to two test specimens validated the preferable accuracy and robustness of the proposed DL-based approach, and demonstrated its high potential for use in post-earthquake performance assessment of buildings.

show abstract

Section: F I G U R E 1 Proposed Evaluation Approach For the Mechanica...mentioning

confidence: 99%

Deep learning‐based evaluation for mechanical property degradation of seismically damaged RC columns

Miao

et al. 2022

Earthq Engng Struct Dyn

View full text Add to dashboard Cite

show abstract

“…Certain strategies-for example, a combination of CNN with features of point clouds-were carried out to identify the damaged areas through two classes of damaged and undamaged utilizing aerial photos. Another strategy of using DL in damage recognition could be employed by adding a field photograph to the aerial images to increase the accuracy of the damaged classes [18]. A combination of multiscale segmentation with CNN is another approach to increase the classification accuracy of recognizing two classes of damaged and undamaged [13].…”

Section: Of 22mentioning

confidence: 99%

Assessment of Convolutional Neural Network Architectures for Earthquake-Induced Building Damage Detection based on Pre- and Post-Event Orthophoto Images

et al. 2020

View full text Add to dashboard Cite

In recent years, remote-sensing (RS) technologies have been used together with image processing and traditional techniques in various disaster-related works. Among these is detecting building damage from orthophoto imagery that was inflicted by earthquakes. Automatic and visual techniques are considered as typical methods to produce building damage maps using RS images. The visual technique, however, is time-consuming due to manual sampling. The automatic method is able to detect the damaged building by extracting the defect features. However, various design methods and widely changing real-world conditions, such as shadow and light changes, cause challenges to the extensive appointing of automatic methods. As a potential solution for such challenges, this research proposes the adaption of deep learning (DL), specifically convolutional neural networks (CNN), which has a high ability to learn features automatically, to identify damaged buildings from pre- and post-event RS imageries. Since RS data revolves around imagery, CNNs can arguably be most effective at automatically discovering relevant features, avoiding the need for feature engineering based on expert knowledge. In this work, we focus on RS imageries from orthophoto imageries for damaged-building detection, specifically for (i) background, (ii) no damage, (iii) minor damage, and (iv) debris classifications. The gist is to uncover the CNN architecture that will work best for this purpose. To this end, three CNN models, namely the twin model, fusion model, and composite model, are applied to the pre- and post-orthophoto imageries collected from the 2016 Kumamoto earthquake, Japan. The robustness of the models was evaluated using four evaluation metrics, namely overall accuracy (OA), producer accuracy (PA), user accuracy (UA), and F1 score. According to the obtained results, the twin model achieved higher accuracy (OA = 76.86%; F1 score = 0.761) compare to the fusion model (OA = 72.27%; F1 score = 0.714) and composite (OA = 69.24%; F1 score = 0.682) models.

show abstract

“…To solve this problem, an improved version of ANNs called convolutional neural networks (CNNs) has been suggested. The convolutional layer capability of this structure makes it an outstanding model for image processing among various image recantation tasks [16][17][18][19][20][21]. Krizhevsky et al [22] presented AlexNet with five convolutional layers and three fully connected layers.…”

Section: Introductionmentioning

confidence: 99%

Damaged Building Extraction Using Modified Mask R-CNN Model Using Post-Event Aerial Images of the 2016 Kumamoto Earthquake

2022

View full text Add to dashboard Cite

Remote sensing is an effective method of evaluating building damage after a large-scale natural disaster, such as an earthquake or a typhoon. In recent years, with the development of computer vision technology, deep learning algorithms have been used for damage assessment from aerial images. In April 2016, a series of earthquakes hit the Kyushu region, Japan, and caused severe damage in the Kumamoto and Oita Prefectures. Numerous buildings collapsed because of the strong and continuous shaking. In this study, a deep learning model called Mask R-CNN was modified to extract residential buildings and estimate their damage levels from post-event aerial images. Our Mask R-CNN model employs an improved feature pyramid network and online hard example mining. Furthermore, a non-maximum suppression algorithm across multiple classes was also applied to improve prediction. The aerial images captured on 29 April 2016 (two weeks after the main shock) in Mashiki Town, Kumamoto Prefecture, were used as the training and test sets. Compared with the field survey results, our model achieved approximately 95% accuracy for building extraction and over 92% accuracy for the detection of severely damaged buildings. The overall classification accuracy for the four damage classes was approximately 88%, demonstrating acceptable performance.

show abstract

Recognition of Damaged Building Using Deep Learning Based on Aerial and Local Photos Taken After the 1995 Kobe Earthquake

Abstract: A quick and accurate method for investigating damage to buildings is required for proper disaster response. In this study, we investigated whether building damage can be grasped by applying Deep Learning, which is one of machine learning methods, to aerial and local photo

Cited by 10 publications

References 3 publications

Deep learning‐based evaluation for mechanical property degradation of seismically damaged RC columns

Deep learning‐based evaluation for mechanical property degradation of seismically damaged RC columns

Assessment of Convolutional Neural Network Architectures for Earthquake-Induced Building Damage Detection based on Pre- and Post-Event Orthophoto Images

Damaged Building Extraction Using Modified Mask R-CNN Model Using Post-Event Aerial Images of the 2016 Kumamoto Earthquake

Contact Info

Product

Resources

About