About the Project
Massey School of Food and Advanced Technology
New Zealand, Palmerston North
Fields of expertise
Food Science, Spectroscopy, Machine Learning & Artificial Intelligence
Chief Supervisor: Prof Richard Archer
For more details visit
Honey is laid down in cells as comb inside frames, each cell sealed with wax. At cell level most nectar may come from one species of flower. But across a whole frame, different cells may contain honey with less clear floral dominance. This project will develop a method of predicting floral origin of each cell of honey based on capture and processing of hyperspectral images of a fresh honey frame.
We know already that Hyperspectral / Imaging spectroscopy techniques can predict the likely UMF value of Manuka honey. And we know that Manuka honey uniquely contains polyphenolics with characteristic fluorescence signatures. This knowledge gives confidence that adequate information is available within the spectral range of most hyperspectral cameras.
The project may contain the following elements:
Predictions using known honey samples under laboratory conditions.
Several hundred known honey samples are available for inspection using several hyperspectral cameras to develop efficient predictions based on few spectral lines. Different data mining techniques may be explored to develop robust prediction models.
Application of predictions to whole frames in the laboratory setting
Image processing techniques will be needed to isolate pixels for an individual cell, a task supported by a post-doctoral researcher also working in the project. The PhD will determine how best to use data from pixels allocated to a cell.
Development of a prototype for use in the industrial setting
The PhD researcher will assist the post-doctoral fellow to develop an industrial unit applying the predictive tools developed. The PhD would address more fundamental issues and the post-doc address more applied ones.
The phase addresses how best to use the hyperspectral camera to capture the fluorescence signature cell by cell. This will require special illumination and may need a sequence of illuminations and image captures, thus introducing a temporal element to the technique.
Required Academic qualifications
• Master’s degree (or equivalent) in food or process engineering, computational science or similar discipline with a focus on applied mathematics and computing.
• Familiarity with at least one programming language (e.g., R, Matlab Python).
• Experience in data and image analysis, ideally machine vision techniques.
Required personal skills
• Ability for independent work displaying initiative and careful thought
• Analytical and academic approach to research questions
• Good collaborative/social skills
• Proficiency in English, both written and spoken
The successful candidate will receive a FIET scholarship covering fees and stipend for three years.
How to Apply?
Please send your CV and cover letter in a single PDF to Dr Reddy Pullanagari,
[Email Address Removed] with subject ““Characterising honey composition by hyperspectral analysis PhD position”.
Support and Location
Based on your current searches we recommend the following search filters.
Based on our current search criteria we thought you might be interested in these.