The accuracy of Fourier transform profilometry (FTP) is greatly influenced by the aliasing between foundational spectra and zero and higher order spectra. We present a novel method in which a bicolor sinusoidal fringe pattern that consists of two interlaced RGB format base color fringe patterns with π phase difference is projected onto the object through a digital light projector and the deformed color pattern is captured by a color digital camera. Then the deformed color fringe pattern is decoded into two individual fringe patterns with π phase difference using a color-separating technique. After calibration of means and contrast, we subtract one of the two fringe patterns from the other to eliminate the aliasing between foundational spectra and zero and higher order spectra. We provide the calibration algorithm of means and contrast. Computer simulation and experiment verify that this method has an obvious advantage. Compared with the π-phase-shifting technique, only one fringe pattern is required to remove the zero spectra and raise the measurable slope of height variation nearly three times, and no phase-shifting device is required in the experimental setup.