The Use of the Wavelet Transform to Extract Additional Information on Surface Quality from Optical Profilometers

This chapter investigates the use of advanced signal processing techniques especially wavelet transforms to extract additional information from a two dimensional surface profile. The wavelet transform is able to aid the user in quickly assessing, visually, if the surface profile has a periodic or non-periodic component as well as if the profile signal is stationary or non-stationary. In addition, thresholds could be set at different frequencies of interest to automatically determine for the user if a periodic signal is present and if its magnitude is acceptable or not. The basis of this chapter is a doctoral dissertation by Lemaster (2004). A laser based, non-contact profilometer was used for all the surface profiles presented in this chapter though contact profilometers could also benefit from this type of analysis. The original work was conducted for wood and wood-based composites; however the signal processing techniques discussed in this chapter are applicable to all types of surfaces. In fact, an industry that would also like to determine if a surface profile is stationary or not or has periodic components is the road surface industry. They routinely use laser based optical profilometers very similar to the type used in this study except for the optics used to obtain the desired range and sensitivity. They are interested in detecting and quantifying pot holes, ruts, and washboard which are very similar to the surface characteristics of interest to the wood industry but on a different scale.