High-throughput, algorithmic determination of pore parameters from electron microscopy

•K-means and computer vision algorithm were used to compute the pore parameters.•The platform performs orders of magnitude faster after recording the image acquisition process and even during image acquisition.•The platform performs parallel computing algorithm when the number of samples is large.•The platform can be applied to other high-throughput study of porous materials by modifying the tool parameters. For the determination of magnesium alloys pore parameters in electron microscopy, a powerful and flexible platform was built, which was based on machine learning and parallel algorithms. This allows one to describe porosity, pore diameter and pore spacing accurately and quantitatively. The platform performs orders of magnitude faster after recording the image acquisition process and even during image acquisition. Automated analysis can provide strong support for high-throughput characterization techniques, and improve efficiency, accuracy, and repeatability significantly compared to manual analysis. The algorithm is a promising way to calculate the porosity much more faster and efficient.