Impact-Echo Dataset for Concrete Inspection with Ground Truth Schematics

This dataset contains 1824 Impact-Echo signals collected over concrete with artificial defects installed. Dataset is collected in two parts (1 and 2). For more information please check: "Robotic Inspection and Subsurface Defect Mapping Using Impact-echo and Ground Penetrating Radar" by Ejup Hoxha, Jinglun Feng et al. (10.1109/LRA.2023.3290386) Note: Ground truth is shown in the schematics. Dataset does not have labels, but one can derive labels using schematics and other provided information. Moreover, due to the success of our learning-based Impact-Echo model, one can always consult our published model. Provided python script is a guide on how to load the `.npy` arrays, and plot them. Schematics and slab skeleton are included in a pdf file. Please cite datasets if you use them for your research.