On-site inspectie van voedselveiligheid en kwaliteit met draagbare spectroscopie, imaging en deep learning (KB-33-007-006)

Project: EZproject

Project Details


Reason and purpose of the research

The latest developments in fast diagnostics using optical equipment is the (I) combinatorial approach of different forms of spectroscopics and image analysis by means of for example deep learning and (II) the miniaturization of the (combination of) sensors. In this preliminary work towards more generic and on-site spectroscopics and image analysis for food and animal inspection applications, the competences of the Wageningen Research parts are synchronized:


Aim for RIKILT:

  • Apply image analysis learning.
  • Identify killer application (s) for food safety where hyperspectral information has added value. "High-profile applications".
  • Plan activities with the new ADX team (WFBR, WBVR, WPR (to be added to team in 2019), RIKILT) for 2019 and up.


Goal for WFBR:

  • A general hyperspectral solution (within specified limits) for different applications. Hardware / software including machine learning / deep learning for hyperspectral data.
  • Learning which food safety applications are interesting for hyperspectral images.


Goal for WBVR:

  • Acquiring knowledge on hyperspectral imaging and applications.


Effective start/end date1/01/1831/12/18