[TIMEPIX] Add notebook to compute centroids of pixel events
Description
The Timepix3 detector is currently being commissioned at SQS as a so called "timestamping camera", i.e. a pixel-based detector that also measures the moment in time a particular is illuminated. It will be primarily used for velocity map imaging (VMI) in the AQS and possibly NQS endstation.
Conceptually somewhat similar to REMI but with a different detection scheme, this technique images individual and charged particles impinging on the VMI. As these are amplified via MCPs, the resulting charge cloud will typically light up several pixels over an extended amount of time on the Timepix detector. An important first step is therefore to find suitable centroids to these groups of illuminated pixels, representing the position, intensity and time of the original particle impact. For reference, this happens implicitly at the detector level for REMI when computing positions and time from delay signals.
@bjoerns from SQS provided an initial notebook implementing this centroiding process using scitkit-learn saved for reference in https://git.xfel.eu/detectors/pycalibration/-/commit/f6f42b92e005ea757427b6a8ff13f9a7b269444e. So far I've only adapted this notebook to the xfel-calibrate notebook API and added saving the results in EuXFEL HDF files to disk. It will be tested using the OfflineCalRunner device on ONC in the user beamtime this week to work towards a stable version deployed on Maxwell.
How Has This Been Tested?
Ongoing
Types of changes
- New feature (non-breaking change which adds functionality)