Improvement of GPR-Based Rebar Diameter Estimation Using YOLO-v3

Abstract

As the frequency of earthquakes has increased in Korea in recent years, designing earthquake-resistant facilities has been increasingly emphasized. Structures constructed with rebars are vulnerable to shaking, which reduces their seismic performance and may result in damage to human life and property. Because the construction of facilities requires the maintenance of sub-constructions, such as by cutting rebars or compensating for missing rebars, information on rebar diameter is required. In this study, the YOLO-v3 algorithm, which has the fastest object recognition performance, was applied to the structural correction data, and a basic experiment was conducted in the air to predict the diameter of rebars in a facility, in real time based on ground-penetrating radar data. The reason for using the YOLO-v3 algorithm is that in the case of GPR data that change slightly according to the diameter of the reinforcing bar, it is difficult to discriminate with the naked eye, and the result may change depending on the inspector. The model achieved a higher accuracy than conventional rebar detection and diameter prediction methods. In addition, the possibility of real-time rebar diameter prediction during construction, using the proposed method, was verified.

Two-stage method based on the you only look once framework and image segmentation for crack detection in concrete structures

“…The YOLO framework has been successfully used in several diverse applications of civil engineering, such as pedestrian detection [46], real-time face detection [47], license plate detection [48], spilled load detection on freeways [49], pothole detection [50], traffic load distribution detection [51], worker and heavy construction equipment identification on site [52], building component identification [53], rebar diameter estimation [54], building footprint estimation [55], traffic management [56], pavement distress detection [57], crack detection [58][59][60][61], and maintenance [62][63][64]. The following sections describe the details of the structure of YOLOv5.…”

A Vision-Based Collision Monitoring System for Proximity of Construction Workers to Trucks Enhanced by Posture-Dependent Perception and Truck Bodies’ Occupied Space

“…Among various object detection techniques of region-based CNNs (R-CNNs), YOLO-v3 [29] demonstrates object detection speed that is more than 100 times faster than that of Fast R-CNNs [30], thanks to two model structure updates from YOLO-v1 [31]. In addition, the YOLO-v3 model has the advantage of being able to learn universal objects well, so it has been recently adopted in various object recognition models in the construction field [32][33][34].…”

Research on Detection Method for Tunnel Lining Defects Based on DCAM-YOLOv5 in GPR B-Scan

“…Kim et al [17] proposed a road defect recognition method based on a convolutional neural network (CNN), which uses CNN to locate road defects after GPR image thresholding. Park [18] studied the performance of the YOLOv3 algorithm in real-time prediction of rebar diameters in facilities, and the result showed that the method can achieve real-time prediction. Yang et al [19] showed that YOLOv5l can achieve the highest detection accuracy and effectively detected the coal fire range, providing a basis for coal fire disaster control.…”