An Automated 2-D Line-Shift Measurement From Smoothed and Leveled Diagnostic Interferometric Images of Exploded-Wire Plasma

A novel, automated, 2-D line-shift measurement algorithm is presented for optical interferometry in plasma diagnostics. By using the smoothing and leveling (SL) algorithm as a preprocessing stage to the Fourier transform method (FTM), the proposed SL-FTM algorithm extracts the line shift without a priori knowledge of the spectral properties of the image, a common requirement of other FTM-based algorithms. The algorithm is simple to implement, and demonstrated for side-on views of plasma from exploded wires, where the interference patterns suffer from low contrast, low signal-to-noise ratio (SNR), and spatially varying intensity. SL-FTM and FTM are compared via Monte Carlo simulation of noisy images with realistic background variation. They are shown to have the accuracies of 0.019 and 0.016 lines, respectively. In addition, the cause of the accuracy difference is studied using a modified Jaccard similarity measure. The measure shows how SL-FTM provides a smoother phase surface than FTM, but underestimates the maximum line shift by up to 15%. Also, the well-known contour tracing method (CTM) is automated as SL-CTM. The automated forms of FTM and CTM (i.e., SL-FTM and SL-CTM) permit, for the first time, their direct comparison for a wide range of noisy images. SL-CTM, and by extension, CTM, achieves the best accuracy of 0.013 lines. Finally, each method is shown to have accuracy that exceeds the standard diagnostic accuracy requirement of 0.05 lines.