Development of a Neural Network Embedding for Quantifying Crack Pattern Similarity in Masonry Structures (original) (raw)
Related papers
International Journal of Advanced Computer Science and Applications
It is critical to develop a method for detecting cracks in historic building concrete structures. This is due to the fact that it is a method of preserving historic building and protecting visitors from the collapse of a historic structure. The purpose of this research is to determine the best method for identifying cracks in the concrete surface of old buildings by using cracked images of old buildings. The various surface textures, crack irregularities, and background complexity that distinguish crack detection from other forms of image detection research present challenges in crack detection of old buildings. This study presents a framework for detecting concrete cracks in old buildings in Semarang's old town using a modified Convolutional Neural Network with a combination of several convolutional layers. This study employs ten convolutional layers (Deca Convolutional Layer Neural Network (DCL-NN)) to provide mapping features for images of concrete cracks in ancient buildings at preservation area. This study also compares commonly used machine learning models such as KNeighbors (n neighbors=3), Random Forest, Support Vector Machine (SVM), ExtraTrees (n estimators=10), and other CNN-pretained models such as VGG19, Xception, and MobileNet. Four performance indicators are used to validate each model's performance: accuracy, recall, precision, F1-score, Matthews Correlation Coefficient (MCC), and Cohen Kappa (CK). This study's data set is comprised of primary data obtained from cracked and normal images of several buildings in Semarang's old town. The accuracy of this study using DCL-NN is 98.87%, recall is 99.40%, precision is 98.33%, F1 is 98.86%, MCC is 97.74%, and CK is 98.86% for crack class. From this study, it was found that the ten convolution layers have higher classification performance compared to other comparison models such as machine learning and other CNN models and are more effective in detecting cracks in concrete structures.
Modeling of masonry failure surface under biaxial compressive stress using Neural Networks
Construction and Building Materials, 2014
Neural Networks are used in order to approximate the experimental results for masonry failure. A two-step procedure is proposed, with the training of two types of Neural Networks. The NNs showed great performance in fitting the experimental input data. The curves generated by the NNs are continuous and smooth, but not necessarily convex.
Proceedings of the International Symposium on Automation and Robotics in Construction (IAARC), 2018
Masonry structures in historical sites are deteriorating due to ageing and man-made activities. Regular inspection and maintenance work is required to ensure the structural integrity of historic structures. The inspection work is typically carried out by visual inspection, which is costly and laborious, and yields to subjective results. In this study, an automatic image-based crack detection system for masonry structures is proposed to aid the inspection procedure. Previous crack detection systems generally involve the extraction of hand crafted features, which are classified by classification algorithms. Such approach relies heavily on feature vectors and may fail as some hidden features may not be extracted. In this study, we propose a crack detection system which combines deep Convolutional Neural Networks (CNN) and Support Vector Machines (SVM). CNN is used in extracting features from RGB images and SVM is used as an alternative classifier to a softmax layer to enhance the classification ability. A dataset containing images of cracks from masonry structures was created using a digital camera and an unmanned aerial vehicle from historical sites. The images were used for training and validating the proposed system. It is shown that the combined CNN and SVM model performs better than the model using CNN alone with the detection accuracy of approximately 86% in the validation images. It is also shown that the system can be used to detect cracks automatically for the images of masonry structures, which is useful for inspection of heritage structures.
Performance Evaluation of Neural Networks in Concrete Condition Assessment
Transportation Research Record: Journal of the Transportation Research Board
A neural network modeling approach is used to identify concrete specimens that contain internal cracks. Different types of neural nets are used and their performance is evaluated. Correct classification of the signals received from a cracked specimen could be achieved with an accuracy of 75 percent for the test set and 95 percent for the training set. These recognition rates lead to the correct classification of all the individual test specimens. Although some neural net architectures may show high performance with a particular training data set, their results might be inconsistent. In situations in which the number of data sets is small, consistent performance of a neural network may be achieved by shuffling the training and testing data sets.
STRUCTURAL INTEGRITY AND LIFE, 2021
The current methodology is focused to investigate a crack assessment method for transit mass dynamics problem in the domain of improved Recurrent Neural Networks (RNNs) methodology. A cracked simply support beam under the action of transit mass is considered as a case study for the present analogy. The knowledge-based Elman's RNNs (ERNNs) approach has been implemented in this problem to find out the position and severity of crack on the beam in a supervised mode. The Levenberg-Merquardt's (L-M) back propagation mechanism or algorithm has been applied to train the knowledge based ERNNs structure. To ensure the robustness of the anticipated investigation, a numerical problem is prepared and analysed. The entire crack detection method has been performed in a supervised mode. The results obtained from ERNNs approach are compared with numerical ones and found to be well convergent. Ključne reči • prslina • ERNNs • L-M algoritam sa povratnim prostiranjem greške Izvod Metodologija istraživanja se fokusira na metodi za procenu prsline u problemu dinamičkog prenosa mase, u okviru domena poboljšanih Rekurentnih Neuronskih Mreža (RNN). U ovoj analogiji se razmatra prosta greda sa prslinom pod dejstvom prenosa mase kao studija slučaja. Pristup sa bazom znanja RNN Elmana (ERNN) je implementiran u ovaj problem radi iznalaženja položaja i veličine prsline u nosaču u modu kontrole. Levenberg-Merkartov (L-M) algoritam ili mehanizam sa povratnim prostiranjem greške je primenjen za treniranje baze znanja ERNN strukture. Radi obezbeđenja robustnosti ovih istraživanja, pripremljen je i analiziran numerički problem. Celokupna metoda za otkrivanje prsline je izvedena u modu kontrole. Dobijeni rezultati pristupom ERNN su upoređeni sa numeričkim, gde je poklapanje u dobroj meri konvergentno. Crack detection in structure by improved recurrent neural Otkrivanje prsline u konstrukciji poboljšanim pristupom
Frontiers in Materials, 2023
The degradation of infrastructures such as bridges, highways, buildings, and dams has been accelerated due to environmental and loading consequences. The most popular method for inspecting existing concrete structures has been visual inspection. Inspectors assess defects visually based on their engineering expertise, competence, and experience. This method, however, is subjective, tiresome, inefficient, and constrained by the requirement for access to multiple components of complex structures. The angle, width, and length of the crack allow us to figure out the cause of the propagation and extent of the damage, and rehabilitation can be suggested based on them. This research proposes an algorithm based on a pre-trained convolutional neural network (CNN) and image processing (IP) to obtain the crack angle, width, endpoint length, and actual path length in a concrete structure. The results show low relative errors of 2.19%, 14.88%, and 1.11%, respectively for the crack angle, width, and endpoint length from the CNN and IP methods developed in this research. The actual path length is found to be 14.69% greater than the crack endpoint length. When calculating the crack length, it is crucial to consider its irregular shape and the likelihood that its actual path length will be greater than the direct distance between the endpoints. This study suggests measurement methods that precisely consider the crack shape to estimate its actual path length.
Masonry Compressive Strength Prediction Using Artificial Neural Networks
Communications in Computer and Information Science, 2019
The masonry is not only included among the oldest building materials, but it is also the most widely used material due to its simple construction and low cost compared to the other modern building materials. Nevertheless, there is not yet a robust quantitative method, available in the literature, which can reliably predict its strength, based on the geometrical and mechanical characteristics of its components. This limitation is due to the highly nonlinear relation between the compressive strength of masonry and the geometrical and mechanical properties of the components of the masonry. In this paper, the application of artificial neural networks for predicting the compressive strength of masonry has been investigated. Specifically, back-propagation neural network models have been used for predicting the compressive strength of masonry prism based on experimental data available in the literature. The comparison of the derived results with the experimental findings demonstrates the ability of artificial neural networks to approximate the compressive strength of masonry walls in a reliable and robust manner.