An image segmentation algorithm for the detection of rebar in bridge decks from GPR scans

Ground penetrating radar (GPR) scans of rebar embedded in concrete structures produce distinct hyperbolic signatures. These signatures can be removed from the data as a preprocess to the identification of other embedded objects. This work is a step towards achieving an unsupervised detection of defects in concrete bridge decks. The proposed algorithm detects the signature of rebar(s) in a GPR scan through image segmentation, arc detection, and a thresholding process. The position and depth of the rebar(s) and the dielectric properties of the concrete are derived.