A comparison of some methods for direct 2D reconstruction from discrete projected views

Preeti Gopal, Ajit Rajwade, Sharat Chandran, Imants Svalbe

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review


Tomographic acquisitions can be described mathematically as discrete projective transforms. Direct reconstruction methods aim to compute an accurate inverse for such transforms. We assemble a limited set of measurements and then apply the inversion to obtain a high-fidelity image of the original object. In this work, we compare the following direct inversion techniques for sets of discrete projections: Radon-i(inverse)Radon, a least squared error method and filtered back-projection for Mojette inversion. We observe that filtered back-projection is the best of these methods, as the reconstruction errors that arise using this method depend least strongly on the image structure. We aim to improve results for the filtered back-projection method by optimizing the design of the regularizing filter and here present work towards eliminating the regularization threshold that is used as part of this method.
Original languageEnglish
Title of host publicationDiscrete Geometry for Computer Imagery
Subtitle of host publication19th IAPR International Conference, DGCI 2016, Nantes, France, April 18–20, 2016, Proceedings
EditorsNicolas Normand, Jeanpierre Guédon, Florent Autrusseau
Number of pages12
ISBN (Electronic)9783319323602
ISBN (Print)9783319323596
Publication statusPublished - 1 Jan 2016
EventInternational Conference on Discrete Geometry for Computer Imagery 2016 - La Cité Nantes Events Center, Nantes, France
Duration: 18 Apr 201620 Apr 2016
Conference number: 19th
https://link.springer.com/book/10.1007/978-3-319-32360-2 (Proceedings)

Publication series

NameLecture Notes in Computer Science
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


ConferenceInternational Conference on Discrete Geometry for Computer Imagery 2016
Abbreviated titleDGCI 2016
OtherDiscrete Geometry for Computer Imagery
19th IAPR International Conference, DGCI 2016, Nantes, France, April 18-20, 2016. Proceedings
Internet address

Cite this