Biologically inspired high dynamic range imaging for use in machine vision /

Date

2017

Authors

Griffiths, Daniel,

Editors

Advisors

Journal Title

Journal ISSN

Volume Title

Type:

thesis

Citation

Statement of Responsibility

Conference Name

Abstract

This thesis offers contributions to the HDR imaging pipeline. These contributions are designed for machine vision applications, prioritising objective measures of signal quality, feature detectability and simplicity of implementation. A novel, Signal to Noise Ratio (SNR) derived model, for weighting traditional image stacks for HDR estimation is presented and empirically shown to improve signal quality by up to 30dB compared to the current state-of-the-art. This benefit is found primarily in the region of low irradiance, where signal quality is typically lowest.

School/Discipline

University of South Australia. School of Engineering
School of Engineering

Dissertation Note

Thesis (PhD(Mechanical and Industry Engineering))--University of South Australia, 2017.

Provenance

Copyright 2017 Daniel Griffiths

Description

1 ethesis (xviii, 236 pages) :
colour illustrations, colour charts
Includes bibliographical references (pages 211-236)

Access Status

506 0#$fstar $2Unrestricted online access

Rights

License

Grant ID

Published Version

Call number

Persistent link to this record