Holographic vision system based on non-diffractive optical scanning holography and deep learning

Research output: Conference Papers (RGC: 31A, 31B, 32, 33)31A_Invited conference paper (refereed items)

View graph of relations

Related Research Unit(s)

Detail(s)

Original languageEnglish
Publication statusPublished - 18 Nov 2019

Conference

TitleSPIE/COS Photonics Asia 2019
LocationHangzhou International Expo Center
PlaceChina
CityHangzhou
Period20 - 23 October 2019

Abstract

We proposed a holographic vision system (HVS) based on a hologram acquisition and a hologram classification units, for capturing and identifying holograms of deformable objects. Non-diffractive optical scanning holography (ND-OSH) is used to capture holograms of physical objects, and a deep learning classifier is applied to deduce the identity of the hologram. Our proposed HVS is evaluated with the set of handwritten numerals. Experimental results reveal that with our proposed HVS, holograms of the test samples can be captured, and subsequently classified with accuracy of over 99.5%.

Research Area(s)

  • Convolutional neural network, Deep learning, Hologram classification, Non-diffractive optical scanning holography, Optical Scanning holography

Bibliographic Note

Research Unit(s) information for this publication is provided by the author(s) concerned.