Near-surface velocity models are important for deep imaging of mineral deposits with seismic exploration. The near-surface can be quite complex from loose, highly heterogeneous materials to stiff, fractured rocks. Surface-wave analysis can be an effective method to image the shallow subsurface of such challenging environments. Here, we propose a workflow that includes several processing and inversion steps. Initially, for the optimization of the processing parameters, we assess the presence of sharp lateral variations with a method based on the measured energy of Rayleigh waves. Then, using a moving window of receivers, we extract Rayleigh-wave dispersion curves along the acquisition line as the maxima of the f-k spectrum. Finally, the dispersion curves are inverted using a laterally constrained inversion scheme. The proposed methodology has been tested on legacy data from a mining field.