Automatic Concrete Damage Recognition Using Multi-Level Attention Convolutional Neural Network.

Hyun Kyu Shin,Ha Young Kim,Yong Han Ahn,Sang Hyo Lee

doi:10.3390/ma13235549

Hyun Kyu Shin, Ha Young Kim + Show 2 more

Open Access

https://doi.org/10.3390/ma13235549

Copy DOI

Abstract

There has been an increase in the deterioration of buildings and infrastructure in dense urban regions, and several defects in the structures are being exposed. To ensure the effective diagnosis of building conditions, vision-based automatic damage recognition techniques have been developed. However, conventional image processing techniques have some limitations in real-world situations owing to their manual feature extraction approach. To overcome these limitations, a convolutional neural network-based image recognition technique was adopted in this study, and a convolution-based concrete multi-damage recognition neural network (CMDnet) was developed. The image datasets consisted of 1981 types of concrete surface damages, including surface cracks, rebar exposure and delamination, as well as intact. Furthermore, it was experimentally demonstrated that the proposed model could accurately classify the damage types. The results obtained in this study reveal that the proposed model can recognize the different damage types from digital images of the surfaces of concrete structures. The trained CMDnet demonstrated a damage-detection accuracy of 98.9%. Moreover, the proposed model could be applied in automatic damage detection networks to achieve superior performance with regard to concrete surface damage detection and recognition, as well as accelerating efficient damage identification during the diagnosis of deteriorating structures used in civil engineering applications.

Highlights

Old buildings and infrastructure in dense urban regions have been consistently exposed to aggressive environmental conditions
This paper proposes a method of analyzing images containing various types of damages and describes the empirical experiments conducted for concrete surface damage identification using convolutional neural network (CNN)
Applied totoanalyze aamulti-damage dataset in order the input input thisstudy, study, CNNwas was applied analyze multi-damage dataset to classify the images into five categories and images into five categories and to images into five categories and to construct a convolution-based concrete multi-damage recognition neural network (CMDnet)

Summary

Introduction

Old buildings and infrastructure in dense urban regions have been consistently exposed to aggressive environmental conditions. These deteriorated components degrade the structural performance, simultaneously revealing several types of defects, such as cracks, delamination, and rebar exposure, on the surfaces of concrete structures. The current visual inspection approach for investigating the conditions of buildings, which is conducted manually by inspectors, is extremely costly and labor-intensive. It is time-consuming to detect damage and determine the type of damage that has occurred [1,2,3,4,5,6]

Methods

Results

Conclusion