Over the course of the Portwings Project, we have gathered a lot of data. Most of this is PTV data. Not all of the data has been analyzed yet. In part, this is because processing the data properly is not easy. Our data is inherently noisy and we do not fully understand what is wrong with it yet. In addition, the DaVis software does not offer a lot of insight into the Shake-the-box algorithm (STB). It is a bit of a black box where data goes in and results come out.
Open-source variants of the STB algorithm exist, which may alleviate this problem. The goal of this internship is to:
1. Identify what factors in raw data influence the result of the particle tracking algorithm.
2. How these factors can be made measurable.
3. How data preprocessing and the PTV algorithm itself can be optimized to mitigate the effects of these factors.
The learning goals of the internship can be summarized in the following points:
1. Distinguish standard PTV tracking inaccuracy, coming from the experiment itself, and the error due to software inaccuracy.
2. Propose methods of quantification of the different types of inaccuracy.
3. Be able to adapt software and algorithms to reduce inaccuracy.