Phase Unwrapping By Neural Network

Case ID:
UA19-061
Description:

Title:  Phase Unwrapping By Neural Network

 

Invention:  This document in-part describes systems and methods that can be used to facilitate reliable and accurate phase unwrapping. The techniques that can be implemented to obtain unwrapped phase using segmentation approaches.  The techniques are also robust against noises in the input data, thereby providing fast and reliable unwrapped phase results.

 

Background:  Phase unwrapping is an important step in interferometry to remove the ambiguity in the interferograms to obtain the correct phase.  Fringe projection techniques are a commonly used three-dimensional (3D) measurement method for non-specular surfaces, by projecting a series of fringes on the object surfaces.  Phase can be obtained from a series of deformed fringes. Phase unwrapping is needed to process  the wrapped phase in order to reconstruct object surface.  Conventional phase unwrapping approaches are time-consuming and noise sensitive.

 

 Advantages:

*  Fast and reliable

*  Much less sensitive to noise

 

Application:

*  Wavefront sensing

*  Optical metrology

 

 

Patent Information:
For Information, Contact:
Kennedy Nyairo
Sr. Licensing Manager, Optical Sciences
The University of Arizona
knyairo@email.arizona.edu
Inventors:
Rongguang Liang
Junchao Zhang
Xiaobo Tian
Jianbo Shao
Keywords:
© 2019. All Rights Reserved. Powered by Inteum