Computer vision approaches for the absolute energy calibration of Bragg crystal setups

Abstract

Bragg crystals are widely used in free electron lasers to diffract and isolate an extremely narrow spectral range from the initial self-amplified spontaneous emission (SASE) signal. By rotating the crystal along one of its axes, multiple reflections can be generated as per Bragg’s law. A measurement model based on Bragg’s law was established at the SASE2 beamline of the European XFEL after extensive hours of hard x-ray self-seeding (HXRSS) operation. This model characterizes the crystal’s range of motion and reflections. In this work, two computer vision techniques, the Hough transform and template matching, are implemented to identify the absolute photon energy from photon diagnostic images. By comparing the spectrometer correlation scans with the model, the actual operational energy can be determined from the measured data using these techniques, enabling accurate calibration as well as the ability to freely scan the photon energy with a self-seeded beam over a wide energy range.