Three-dimensional target recognition using the phase information from a Fourier transform profilometer

Abstract

We propose a new technique for three-dimensional (3D) target recognition using the phase information from a Fourier transform profilometer. Instead of cross-correlating the 3D target/nontargets induced distorted grating patterns or the associated phase differences directly, the phase information are rather utilized to synthesize two complex harmonic functions, spatial frequency distributions of which are proportional to the computed profilometric phase maps of the target and the nontargets and hence to respective shape information. These complex harmonic functions due to the target and the nontargets are cross-correlated to produce highest correlation peak at the target location and almost no correlation peak for others. The feasibility of the technique is demonstrated by simulation and experiments showing an excellent discrimination capability for 3D target recognition.