Recognising copywrited images in open educational resources

Primary Challenge Owner: ROCVA and SURF
Secondary Challenge Owner(s): none

SURF and ROCvA both act as Primary Challenge owners, with SURF being responsible for the pilot budget and therefore for approving the budgets in the Pilot Plan (as described in Section 1.3). ROCvA provides the pilot location. Further division of roles will be jointly explained and worked out during the deep dive.


More and more educational institutions are investing in open learning materials. Lecturers as well as students create learning materials themselves and make these widely available. A problem that arises with these Open Educational Resources(OER) is the use of copyrighted images. Usage of such images without permission is not permitted and can lead to heavy fines. The authors of these Open Educational Resources are not always sufficiently aware of this. In addition, auditing these materials by the educational institutions is time-consuming. The CopyRIGHT tool developed by SURF already classifies text from PDFs as copyright, open access or own work. Preferably images should be able to be included in this as well. Particularly now that institutions are increasingly focusing on developing, using and widely sharing OER, respecting copyright is important.

The challenge:

Develop a solution that can recognize copyrighted images in open educational resources to reduce the use of images without permission and to make sure additional charges can be prevented. 


  • Preferably in Dutch, but English is allowed as well
  • The tool is preferably open source
  • The tool would assess single images and preferably also images in PDFs, Powerpoints and Word documents for copyright
  • This tool must be integrated into the CopyRIGHT tool. CopyRIGHT provides integration with LMSs such as CANVAS.
  • ROCvA’s materials are located in Xerte*. A plugin for Xerte is a plus, but could also be developed by CopyyRIGHT.
  • *Xerte is an open source authoring tool and supports HTML5, JavaScript and CSS.
  • There is no database of all copyrighted images to check. It is a challenge to compare images with material that is (easily) findable and available for those who develop learning materials. It should also be investigated whether metadata is available with images that can be used for assessment.


Do you have a possible solution for this challenge? Great! You can apply here. Fill in your email address to get full access to the documents and the application page. Applications will be accepted from April 4th until May 24th 2023.