Real-time Defect Detection Method without Training Data for Hammering Test Support System

As a nondestructive diagnosis method, hammering test is a popular approach for detecting defect on concrete surfaces. However, as the number of concrete structures that require inspection is increasing, it is an important issue in recent years to reduce inspection time and human cost during hammering test. This paper introduces a novel method to detect defect on concrete surfaces in real-time during hammering test. The proposed method has two features; 1) automatically analyzing the acoustic data based on the statistical data of hitting sounds without prior training data, 2) the diagnostic results immediately displayed by projection mapping to the hit positions. We show the effectiveness of the proposed system through the experiments by using test concrete block with defects.