Affiliation:
1. Xi’an Jiaotong University
Abstract
In this paper, an ANLVENet
speckle suppression method
in holographic phase fringe patterns with different level noises is
proposed based on FFDNet,
combined with asymmetric pyramid non-local block with a verge
extraction module. The experimental results are compared to three
network models and several representative algorithms. It is shown that
the ANLVENet method not only has better superiority in the speckle
suppression with different noise levels, but also preserves more
details of the image edge. In addition, another speckle noise model is
applied in the phase fringe patterns to prove the stronger
generalization of the ANLVENet algorithm. The proposed method is
suitable for suppressing the speckle with different levels in a large
noise range under complex environmental conditions.
Funder
National Natural Science Foundation of
China
Key Research and Development Plan Project
in Shaanxi Province
Natural Science Basic Research Program of
Shaanxi
Xi’an Science and Technology Plan
Project
Seed Fund for Creativity and Innovation
of Postgraduates of Xi’an University of
Technology
Subject
Atomic and Molecular Physics, and Optics,Engineering (miscellaneous),Electrical and Electronic Engineering