Artificial intelligence-based internal defect detection system for concrete members

An artificial intelligence-based internal defect detection system for concrete members comprising:

• a signal detection unit (3), which includes a first detection unit (31) attached or placed close to the concrete member (1) and detecting a reflection signal by excitation applied to a concrete member, which is a target inspection, to measure both the reflection signal and external noise simultaneously, and a second detection unit (32) installed to be spaced apart from the concrete member (1) at a position close to the concrete member (1) to measure external noise;
• and
• a data processing unit (4), which includes a signal processing unit (41) which generates an impact echo signal by calculating a difference between a detection signal of the first detection unit (31) and a detection signal of the second detection unit (32), generates a frequency signal from the impact echo signal, and extracts instantaneous frequency data from the frequency signal, and a defect determination module (42) which receives the instantaneous frequency data generated by the signal processing unit (41) as feature information to determine occurrence of defects of the concrete member (1), builds a defect occurrence decision model using a bidirectional LSTM model, and determines defects of the concrete member (1) using the defect occurrence decision model based on the detected signal.