Adaptive optics devices in astronomy compensate the distortions introduced by the atmosphere in the quality and resolution of images taken by ground-based telescopes. The proper knowledge of the spatial and temporal behaviour of the atmospheric turbulence is crutial for optimizing the design of multiconjugate adaptive optics systems to satisfy the excellent image quality requirements of the new generation of large and extremelly large telescopes. Atmospheric turbulence monitoring is being carried out at astronomical sites using different techniques, being generalized SCIDAR one of the most popular. Generalized SCIDAR technique provides enough information to derive the vertical structure of the atmospheric turbulence (C2N(h)) as well as the velocity of the turbulent layers (Vtur(h)). In this work, we present a new and automatic method to derive Vtur(h) from generalized SCIDAR data based on wavelet analysis.