Artificial intelligence-based Spectral Correction for Hyperspectral Imaging Sensors

Leuven - Master projects/internships
More than two weeks ago

This project aims to develop AI-based methods for spectral correction for hyperspectral imaging sensors

Hyperspectral cameras enable the collection and processing of information across the electromagnetic spectrum. Thanks to the world leading technology, the IMEC hyperspectral imaging sensor plays an important role in various applications like remote sensing, precision agriculture, health care, and food industry. The manufacturing process of the hyperspectral filters on the sensors can introduce deviations from the original design within certain tolerances, exhibit higher order harmonics and is combined with normal sensors behavior, like crosstalk or wavelength dependent sensitivity (QE) variability. Using a post-manufacturing calibration step, high quality spectra can still be acquired, using a process called spectral correction. However, under certain application and scene dependent conditions, like the usage of specific light types or special lenses, the standard spectral correction can introduce some unwanted artefacts. 

Based on the underlying physics of imec’s hyperspectral imaging technology, we are now looking for AI-based solutions for spectral correction that would be optimized to cope with a range of illumination conditions or scene/application specific conditions. The goal of this project would be to develop one or more reliable AI-based methods for spectral correction under different experimental conditions. Starting from the Neural Network method, the candidate will have the freedom to extend this study to deep learning and different kinds of machine learning methods which might be a solution for this challenge.

What we offer:

  • Work with a world leading hyperspectral camera system in an experimental environment.
  • Real data from the lab and projects on different applications.
  • Technical support from a multi-disciplinary team with expertise from hardware design to algorithm development.


  • Literature review (10%)
  • Experiment and data collection (20%)
  • Data preprocessing and processing (10%)
  • Model development (50%)
  • Reporting and presentation (10%)

Who you are?

  • Knowledge of machine learning algorithms.
  • Knowledge of data processing tools, like matlab, or python.
  • Motivated student eager to work independently and expand knowledge in the field.
  • Good level of English.

Type of Project: Thesis or internship project

Duration: 3-6 months

Master program: Electrical Engineering or Computer Science

Supervising scientist: For further information or for application, please contact Yuqian Li (

Share this on


This website uses cookies for analytics purposes only without any commercial intent. Find out more here. Our privacy statement can be found here. Some content (videos, iframes, forms,...) on this website will only appear when you have accepted the cookies.

Accept cookies