Skip to content
ST
Research

Computer Vision Pipeline for LPBF Spatter Tracking

Developed a computer-vision and ML pipeline for detecting, tracking, and measuring spatter ejections in high-speed LPBF imaging.

01 / Overview

I developed a computer-vision and ML-based analysis pipeline for laser powder bed fusion research. The project focused on detecting and tracking spatter ejections from high-speed imaging data and extracting measurable features such as velocity, size, and ejection angle.

02 / Technical focus

This project combined computer vision, machine learning, experimental data analysis, and materials and manufacturing research. It helped me build experience with noisy real-world research data, high-speed imaging, feature extraction, and reproducible Python analysis workflows.

03 / What I worked on

  • Built a Python computer-vision pipeline to detect and track spatter ejections in LPBF high-speed imaging experiments.
  • Processed 30,000 fps imaging data to support quantitative analysis of spatter behavior.
  • Used OpenCV and scikit-learn to improve tracking accuracy by approximately 20% over manual methods.
  • Applied feature extraction and statistical modeling to analyze spatter velocity, size, and ejection angle.
  • Visualized experiment outputs using Matplotlib to support reproducible analysis and research review.
  • Connected computer vision outputs to process-level interpretation for additive manufacturing experiments.

04 / Tools and topics

PythonNumPyPandasSciPyOpenCVscikit-learnMatplotlibComputer visionFeature extractionStatistical modelingHigh-speed imagingLPBFReproducible data analysis