Automatic Surface Flatness Control using Terrestrial Laser Scanning Data and the 2D Continuous Wavelet Transform

Evaluating flatness defects in surfaces during the construction of a building can prevent problems during subsequent construction tasks or the operation phase. An area of particular interest is that of the flatness of warehouse slabs on which pallet-transporting equipment is operated (nowadays including robots). Flatness control has typically been done with extensively manual methods, such as the Straightedge, F-Numbers or Waviness Index methods. However, these are time-consuming, based on very sparse measurements, and conduct assessment in linear ways only (i.e. in 1D along survey lines, as opposed in 2D on the entire surface), which can lead to inaccurate assessment. This paper presents a novel approach that takes advantage of (1) Terrestrial Laser Scanners (TLS) to speed up data acquisition and provide precise and dense 3D measurements of surfaces; and (2) the 2D Continuous Wavelet Transform (CWT) to deliver a 2D wavelength analysis of surface data with high resolution both spatially and in the frequency domain. The value of the proposed method over existing approaches is discussed and demonstrated with a first experiment conducted with a real concrete slab.