A two-dimensional conditional averaging method for velocity estimation immune to the barber pole effect

Methods for analyzing the motion of coherent structures in coarse-grained imaging data often face challenges due to the barber pole effect, arising when elongated and tilted structures move at an angle with respect to their axis of symmetry. This study presents a velocity estimation technique that overcomes the barber pole effect while remaining well-suited for coarse-grained imaging data. The proposed method employs two-dimensional conditional averaging to characterize the typical shape and motion of large-amplitude structures passing through a reference point, followed by a maximum tracking or contouring approach to estimate the average trajectory, from which the velocity is calculated. The method is validated using synthetic datasets, demonstrating its accuracy and potential applicability to gas puff imaging and beam emission spectroscopy measurements in magnetically confined plasmas.