Graduate Thesis Or Dissertation

 

A Comparison of Techniques Used for the Removal of Lens Flare Found in High Dynamic Range Luminance Measurements Public Deposited

Downloadable Content

Download PDF
https://scholar.colorado.edu/concern/graduate_thesis_or_dissertations/8k71nh44r
Abstract
  • The presence of lens flare has been identified as a limitation associated with luminance measurements obtained using High Dynamic Range Imaging (HDRI) technology. Current documentation is lacking on the magnitude of error lens flare has on luminance values as well as the methods for lens flare removal. The purpose of this study is to provide an evaluation of the effectiveness of two existing lens flare removal techniques. The first technique involves the implementation of blind deconvolution during image postprocessing within commonly used HDRI software. The second technique aims to separate and extract lens flare from luminance values which are true to the scene using a high frequency occlusion mask. The results from this study indicated that while visually problematic, the error in luminance values resulting from lens flare effects a limited number of pixels located near the source. Of those pixels exhibiting significant error, current lens flare removal techniques are not sufficient at correcting this error nor do they adequately remove the visual appearance of lens flare found in HDR images. Future correction efforts may be best served by taking a step away from existing methods and working to create a more robust solution through the development of a Convolutional Neural Network (CNN).
Creator
Date Issued
  • 2017
Academic Affiliation
Advisor
Committee Member
Degree Grantor
Commencement Year
Subject
Last Modified
  • 2019-11-18
Resource Type
Rights Statement
Language

Relationships

Items