Open Access3D object recognition through processing of 2D holograms
Author(s) -
Behzad Bordbar,
Haowen Zhou,
Partha P. Banerjee
Publication year - 2019
Publication title -
applied optics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.668
H-Index - 197
eISSN - 2155-3165
pISSN - 1559-128X
DOI - 10.1364/ao.58.00g197
Subject(s) - holography , optics , computer science , correlation , computation , metric (unit) , artificial intelligence , object (grammar) , computer vision , image processing , cognitive neuroscience of visual object recognition , spatial frequency , pattern recognition (psychology) , mathematics , physics , algorithm , image (mathematics) , geometry , operations management , economics
Correlation of two-dimensional digitally recorded holograms is introduced as a novel approach for object recognition without the need for quantitative assessment of the retrieved complex field, based on the fact that a hologram contains the three-dimensional information of the object. Actual objects with different three-dimensional features such as depth and surface roughness are assessed through processing of the correlation of their two-dimensional holograms. Correlation peak values are extracted as a metric to evaluate correlation of three-dimensional objects. The effect of hologram windowing size on correlation of three-dimensional objects is investigated, and improvements in computation time and dynamic range are assessed. Critical figures of merit used for assessment of correlation of images are applied to the correlation of holograms for object recognition.