Photofragment image analysis via pattern recognition
Sergei Manzhos, Hans‐Peter Loock · Review of Scientific Instruments · 2004
An algorithm is presented that solves two problems associated with the analysis of velocity map images, which are used, for example, in the study of photofragmentation or photoionization processes. The first part of the algorithm identifies the center, the ring radii, and distortions of circularity without any a priori knowledge about the image. Derived from the Hough transform, it is highly robust with respect to uneven distributions of intensity, background signals, and realistic distortions of circularity. In the second independent part of the algorithm the image parameters are calculated using an analytical description of the image. Here the velocity profile, branching ratios, and spatial anisotropy parameters are obtained directly from the raw image for any form of the velocity broadening function, i.e., without the necessity for “inversion” of the image.