Design and Control of a Novel 3-D Piezoelectric Scanning Coaxial Optical Microscope System

This article presents a novel 3-D optical measurement system combining an optical microscope (OM) and a
$z$
-axis piezoelectric scanner. The proposed system includes an advanced controller for the piezoelectric scanner and an algorithm to process OM imaging data. The
$z$
-axis piezoelectric scanner uses a back-propagation neural network (BPNN) to control the upward movement of the sample in a stepwise manner. At the point where each step induces a small steady-state error, the OM captures an image in that position. Each 2-D OM image undergoes denoising via connected-component labeling (CCL) before being converted into 3-D data via image processing. The layered images are then stacked in a 3-D arrangement to construct an accurate 3-D image of the sample.