Image manipulation in scholarly publications: are there ways to an automated solution?
| Date | 07 December 2021 |
| Pages | 1184-1198 |
| DOI | https://doi.org/10.1108/JD-06-2021-0113 |
| Published date | 07 December 2021 |
| Subject Matter | Library & information science,Records management & preservation,Document management,Classification & cataloguing,Information behaviour & retrieval,Collection building & management,Scholarly communications/publishing,Information & knowledge management,Information management & governance,Information management,Information & communications technology,Internet |
| Author | Thorsten Stephan Beck |
Image manipulation in scholarly
publications: are there ways to an
automated solution?
Thorsten Stephan Beck
Library and Information Science, Humboldt-Universit€
at zu Berlin, Berlin, Germany
Abstract
Purpose –This paper provides an introduction to research in the field of image forensics and asks whether
advances in the field of algorithm development and digital forensics willfacilitate the examination of images in
the scientific publication process in the near future.
Design/methodology/approach –This study looks at the status quo of image analysis in the peer review
process and evaluates selected articles from the field of Digital Image and Signal Processing that have
addressed the discovery of copy-move, cut-paste and erase-fill manipulations.
Findings –The article focuses on forensic research and shows that, despite numerous efforts, there is still no
applicable tool for the automated detection of image manipulation. Nonetheless, the status quo for examining
images in scientific publications remains visual inspection and will likely remain so for the foreseeable future.
This study summarizes aspects that make automated detection of image manipulation difficult from a forensic
research perspective.
Research limitations/implications –Results of this study underscore the need for a conceptual
reconsideration of the problems involvi ng image manipulatio n with a view toward the ne ed for
interdisciplinary collaboration in conjunction with library and information science (LIS) expertise on
information integrity.
Practical implications –This study not only identifies a number of conceptual challenges but also suggests
areas of action that the scientific community can address in the future.
Originality/value –Image manipulation is often discussed in isolation as a technical challenge. This study
takes a more holistic view of the topic and demonstrates the necessity for a multidisciplinary approach.
Keywords Research, Information literacy, Ethics, Algorithms, Automation, Generation and dissemination of
information, Information science
Paper type Research paper
1. Introduction
Nowadays, thorough evaluation of digitally distributed information is more relevant than
ever. In a media landscape in which information can be disseminated extremely quickly,
images are increasingly imp ortant for the authenticat ion of events. Disinformati on
campaigns and fake news complicate distinguishing reliable information from propaganda.
Moreover, images not only influence risk assessment but also play a crucial role in political
opinion making and may thus be impacting the stability of democracy. Farid notes: “The ease
with which digital media can be altered and manipulated is affecting nearly every corner of
our world: media, politics, law, business, and science”(Farid, 2006, p. 6). The flood of images
shared daily makes it almost impossible to comprehensively verify their authenticity and
integrity, or even their origin. On Facebook alone, around 300 million photos are shared every
day (Marr, 2018). The multiple purposes and intentions associated with the use and recycling
of images present an additional challenge.
Pictures often reflect an idealized or beautified version of reality. In the World Press Photo
Award 2015, around 20% of the pictures in the penultimate round had to be disqualified [1]
JD
78,5
1184
Funding: This research was financially supported by the Humboldt-Elsevier Advanced Data and Text
Centre (https://headt.eu/).
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/0022-0418.htm
Received 10 June 2021
Revised 16 June 2021
11 November 2021
Accepted 13 November 2021
Journal of Documentation
Vol. 78 No. 5, 2022
pp. 1184-1198
© Emerald Publishing Limited
0022-0418
DOI 10.1108/JD-06-2021-0113
To continue reading
Request your trial