https://github.com/abought/ori_openforensicactions

Detect image fraud via recoloring: matplotlib and web tools inspired by ORI advanced forensic actions. Not affiliated with or endorsed by ORI etc.

https://github.com/abought/ori_openforensicactions

Science Score: 13.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
  • .zenodo.json file
  • DOI references
    Found 4 DOI reference(s) in README
  • Academic publication links
  • Academic email domains
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (16.1%) to scientific vocabulary
Last synced: 10 months ago · JSON representation

Repository

Detect image fraud via recoloring: matplotlib and web tools inspired by ORI advanced forensic actions. Not affiliated with or endorsed by ORI etc.

Basic Info
Statistics
  • Stars: 0
  • Watchers: 3
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Created over 11 years ago · Last pushed over 11 years ago
Metadata Files
Readme

README.md

Purpose

This represents the beginnings of a Python port of the ORI Advanced Forensic Actions tools.

The original version was developed as an Adobe Photoshop CS4-CS5 plugin; this effort was motivated by a desire to make these tools usable without software licensing fees. This port is provided without warranty, and was produced without affiliation or endorsement from the Office of Research Integrity or original authors.

Below is an example from a retracted publication. (adapted from DOI: 10.1021/nl400959z, Copyright 2013 American Chemical Society). Applying a high-cycle colormap (high_18) emphasizes contrast differences near the boundaries of two nanoparticle rods.

Nanoparticle TEM image

Status

To date, only the gradient LUT (colormap) data has been extracted. These gradients can be used as part of a larger analysis chain in python/matplotlib, or in a simple web-based tool.

Issue reports, feature requests, and code enhancements (via pull requests) are welcome.

Installation

The user must have Python 2.7 or 3.x installed on their computer. Additional dependencies can be installed using the following command:

pip install -r requirements.txt

MS Windows users may need to download precompiled binary versions from the relevant project websites.

Usage

Using colormaps in python & matplotlib

Gradients can be imported for use in scripts with matplotlib, as part of a larger analysis toolchain. A sample command-line tool is provided in the folder python/, which applies a named gradient to the image specified:

python apply_lut.py image_filename.png --lut COLORMAPNAME

For a list of all recognized colormaps, use

python apply_lut.py -h

Web tool

For quick experimentation, a browser-based tool is also provided. This allows you to load an image from your local hard drive and experiment with different gradients (multiple gradients may be needed to see all relevant features). The modified image can then be saved.

A live demo of the web tool is available.

About the gradients

Low Cycle Gradients feature a slow transition from one color to another. Low cycle gradients

Mid Cycle Gradients transition between several colors over a broad range. Mid cycle gradients

High cycle gradients rapidly swap between multiple colors. These gradients will most amplify small differences in local contrast, but some colors may appear more than once. High cycle gradients

Acknowledgements

Lookup tables were developed by the Office of Research Integrity and released as part of the Advanced Forensic Actions ORI Advanced Forensic Actions tools tool suite.

Colormap data was extracted and converted using the grdtocmap conversion utility, based on code from the RE-lab team.

Further reading

To learn more about how these tools (including gradient LUTs) can be applied for fraud detection, see the description of methods in the affidavit of John W. Kruger, Case No. 14-013099-CZ.

Owner

  • Name: Andy Boughton
  • Login: abought
  • Kind: user
  • Location: United States
  • Company: @statgen

GitHub Events

Total
Last Year

Dependencies

requirements.txt pypi
  • matplotlib *
  • numpy *
  • scipy *