Three-Dimensional Object Recognition using Multiplex Complex Amplitude Information with Support Function

We propose a three-dimensional (3D) object recognition method using multiplex complex amplitude with a support function. 3D shape information of an object is directly encoded into the phase factor of the complex amplitude by the Fourier transform profilometry technique. The support function determined on the basis of the amplitude factor of the complex amplitude can remove unnecessary background or unreliable areas of the object due to low reflectance. For robust object recognition, the reference data is produced by spatial shift multiplexing of complex amplitudes derived from many parallax views. Object recognition is performed by correlation between the multiplex reference data and the target object at a single operation. In the experiment, we confirm 3D object recognition using the multiplex reference data including many objects with various rotation angles