Ultrafast Laser Beam Profile Characterization in the Front-End of the ELI-NP Laser System Using Image Features and Machine Learning

Ultrafast laser systems, implemented at the ELI-NP, require exceptional beam quality and spatial stability due to their femtosecond pulse durations and extremely high peak powers. This work presents a diagnostic and computational framework for analyzing the ELI-NP Front-End beam characteristics, where spatial coherence and precise pulse shaping are essential for reliable amplification and experimental consistency. The methodology integrates classical beam diagnostics with image processing and machine learning tools to evaluate anomalies based on high-resolution beam profile images. We use centroid tracking to monitor pointing fluctuations, statistical intensity analysis to detect energy instabilities, and Sobel-based edge detection to evaluate beam sharpness and extract structural features from the beam image. Geometric parameters such as ellipticity, roundness, and symmetry indicators are extracted and examined over time. The system applies an unsupervised Isolation Forest algorithm to detect subtle or short-lived anomalies, identifying irregularities without relying on predefined thresholds. These diagnostics are supported by visual plots and statistical summaries, offering a clear picture of the beam’s behavior under real operating conditions. Results confirm that this integrated approach effectively captures major and minor beam instabilities, making it a practical tool for continuous monitoring and performance optimization in ultrafast laser systems.