Fringe projection profilometry (FPP) has been widely studied for measurements of three-dimensional (3D) shapes. However, the measurement of moving objects is still a challenging problem associated with commonly used methods. Phase-shifting profilometry (PSP) suffers from motion-induced errors when applied to moving objects, while Fourier transform profilometry (FTP) is sensitive to background light and noise. This paper proposes a novel data fusion approach to improve the measurement of moving objects by alleviating noise-induced errors. Firstly, a set of coarse measurements on the shape of the object are obtained at different time instants by FTP. Then the parameters associated with the motion of the object over these time instants are retrieved from these coarse estimates by applying the iterative closest point (ICP) method. Finally, the parameters are utilized to adaptively fuse the coarse measurements based on their quality. Simulations and experiments verify that the proposed approach can significantly improve the measurement accuracy.