Internship Gwyn

Gwyn van der Wal

Meet Gwyn, a Master's student in Applied Mathematics at Delft University of Technology specializing in stochastics. As part of her internship at SG11, she focuses on data transformations, a crucial aspect of our classification algorithm. Gwyn's recent report delves into the influence of whitening transformations on hyperspectral data of the Delta R Shoe Scanner, providing SG11 with valuable insights and methods for implementation. This contribution marks a significant step in our continuous development of improved classification algorithms.

Internship Gwyn

Background

During my graduation project at Stage Gate 11, I researched the influence of whitening transformations on the hyperspectral data received from the Delta R Shoe Scanner.
The Shoe Scanner scans the outside of the shoe for illicit materials. The hyperspectral data then goes through a number of steps before classifying the substances on the shoe as either a concern or as harmless. My thesis focussed on one step of the pre-processing: whitening.
The whitening transformation transforms data to remove unnecessary information. By removing these structures, the interesting structures can be looked at for better classification. This is why Stage Gate 11 B.V. has employed whitening in the pre-processing of their hyperspectral data. The aim of my work was to gain insight into the whitening transformation and how it influences hyperspectral data.

Method

To gain insight, I created synthetic data and made synthetic scans. The signal-to-noise ratio of a target spectrum was calculated, and Monte Carlo simulations were used to reveal hidden patterns in the data.

Results

First of all, I have researched what a target spectrum looks like before and after whitening, helping SG11 better understand this transformation in various scenarios. Secondly, I determined what the inputted data matrix should look like. Further, I calculated the signal-to-noise ratio of a target spectrum after whitening and researched how this ratio can be improved.
With this research, I have given SG11 insight into the whitening transformation and the methods to employ it, which they can use in their continuous development of better classification algorithms.

Experience

I fondly look back at my time at SG11. I would like to thank the entire staff of SG11, who were all willing to help where possible and open for questions and discussions, and for the good times. Especially Rob Satink, his guidance, support and willingness to share his knowledge of spectroscopy and engage in discussions which have truly shaped my thesis. And now I am proud to say that I have finished my Master of Science in Applied Mathematics.