Interpretation of IE data have been carried out by analyzing IE signals in frequency domain to determine the maximum frequency. However, the current peak frequency method can be inaccurate. The purpose of this research is to introduce features in IE signals that can be used for effective classification and interpretation for bridge deck evaluation through statistical analysis and Naive Bayes classifiers. The dataset contained IE data collected from eight slabs created at Advanced Sensing Technology FAST NDE laboratory (FHWA). A set of statistical features in time domain, normalized peak values, and length of preprocessed signals were used to classify the IE data, statistically. Then, Naive Bayes classifiers was employed to recognize defect area. Finally, the result of statistical classification was compared with frequency approach. The result shows that 19 and 21% of the IE signals collected from the defect area have multiple peaks, respectively. However, 85% of the IE signals collected from the sound set had only one peak. A probability classifier was used to find the relationship between the result of the frequency method and statistical analysis. The result shows that 10% of the IE signals were usable for estimating the thickness in the sound group.
Annotated datasets play a significant role in developing advanced Artificial Intelligence (AI) models that can detect bridge structure defects autonomously. Most defect datasets contain visual images of surface defects; however, subsurface defect data such as delamination which are critical for effective bridge deck evaluations are typically rare or limited to laboratory specimens. Three Non-Destructive Evaluation (NDE) methods (Infrared Thermography (IRT), Impact Echo (IE), and Ground Penetrating Radar (GPR)) were used for concrete delamination detection and reinforcement corrosion detection. The authors have developed a unique NDE dataset, Structural Defect Network 2021 (SDNET2021), which consists of IRT, IE, and GPR data collected from five in-service reinforced concrete bridge decks. A delamination survey map locating the areas, extent and classes of delamination served as the ground truth for annotating IRT, IE and GPR field tests’ data in this study. The IRT were processed to create an ortho-mosaic maps for each deck and were aligned with the ground truth maps using image registration, affine transformation, image binarization, morphological operations, connected components and region props techniques to execute a semi-automatic pixel–wise annotation. Conventional methods such as Fast Fourier transform (FFT)/peak frequency and B-Scan were used for preliminary analysis for the IE and GPR signal data respectively. The quality of NDE data was verified using conventional Image Quality Assessment (IQA) techniques. SDNET2021 dataset consists of 557 delaminated and 1379 sound IE signals, 214,943 delaminated and 448,159 sound GPR signals, and about 1,718,083 delaminated and 2,862,597 sound IRT pixels. SDNET2021 addresses one of the major gaps in benchmarking, developing, training, and testing advanced deep learning models for concrete bridge evaluation by providing a publicly available annotated and validated NDE dataset.
Impact echo (IE) is a non-destructive evaluation method commonly used to detect subsurface delamination in reinforced concrete bridge decks. Existing analysis methods are based on frequency domain which can lead to inaccurate assessments of reinforced concrete bridge decks since they do not consider features of the IE signals in the time domain. The authors propose a new method for IE classification by combining features in the time and the frequency domains. The features used in this study included normalized peak values, energy, power, time of peaks, and signal lengths that were extracted from IE signals after they are preprocessed. We used a dataset containing IE data collected from four in-service bridges, annotated using chain dragging. A support vector machine (SVM) classifier was constructed using combined features to classify IE signals. A 1DCNN with unfiltered IE signals and a two-dimensional CNN using wavelet scalograms (2D representations of unfiltered IE signals) were also used to classify IE signals. The SVM model performed significantly better than the other models, with an accuracy rate, true positive rate, and true negative rate of 97%, 92%, and 98%, respectively. The SVM model also generated more accurate defect maps for all investigated bridges. IE data from the Federal Highway Administration’s InfoBridge website were used to investigate the efficacy of the developed models. The investigation yielded promising results for the proposed SVM model when used for a new set of IE data.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.