| description abstract | The precise determination of reinforcement rebar information of constructed facilities is a significant challenge faced by many practitioners in the areas of structural health monitoring (SHM), facility management (FM), and building maintenance and operation. Practitioners commonly use ground penetrating radar (GPR) to determine the horizontal location, depth, and size of rebar based on the reflections in GPR data. However, due to inherent difficulties within the GPR data, such as the unknown time zero, the existence of strong noise, and the blurry signals, simultaneously and accurately determining all three parameters is a challenging task. To tackle this issue, this paper proposes an integrated approach based on pattern recognition and curve-fitting principles to simultaneously determine rebar’s horizontal location, depth, and size. The presented method has been validated on several scanning trials of a shear wall in an in-service concrete structure. Results of conducting the experiments are promising and reveal that (1) the proposed method can identify the zero offset within the GPR scan, eliminate the noise impact in the GPR signal, and extract critical rebar information from the attenuated signal; (2) by implementing through a real-world application, this integrated method could successfully solve the coupling issue and simultaneously determine rebar’s horizontal location, depth, and size; (3) the rebars’ horizontal locations in the test scenarios are consistent, with mean absolute errors of measured depths and sizes of 6.73% and 6.19%, respectively; and (4) considering a range of each rebar size, all rebars are measured correctly. | |
