A novel phase-coding technique based on Fourier-transform profilometry (FTP) is proposed for three-dimensional object recognition. Two spatially multiplexed grating patterns of different spatial frequencies are projected simultaneously onto the objects-target, and the phase changes in the distorted patterns are detected. An algebraic addition or subtraction of these phase values is utilized to code the two-dimensional plans of the objects-target with spatial harmonic modulations. The phase-coded plans of the objects and the target are cross correlated digitally to yield a high correlation peak at the target location. The 2pi ambiguity of the phase associated with FTP has been resolved in correlation results without recourse to phase unwrapping. Experimental results show an excellent discrimination capability for target recognition.