Here, I describe a method for manually tracking particle motions from high-speed imagery of flume experiments using the MTrackJ plugin (https://imagescience.org/meijering/software/mtrackj/) for the open-source software ImageJ (https://imagej.nih.gov/ij/download.html).
High-speed imagery of sediment transport is captured as a series of high-resolution photos taken by a camera positioned above the experimental flume. The camera is positioned directly perpendicular to the bed surface. A plexiglass "sled’’ window is placed on the surface of the surface of the water. This provides a clear view of the bed, preventing distortion of the images caused by light refraction from fluctuations in the water surface. The flow is sufficiently deep that the plexiglass window does not interfere with the flow at the bed surface.
Sequences of images obtained from high-speed photography can be uploaded as an image stack to the ImageJ program. Then using the MTrackJ plugin the user can manually mark the centroid of each particle as it moves through successive frames. ImageJ provides sub-pixel resolution of the centroid coordinates. The software records the pixel coordinates for each particle throughout the set of consecutive images as individual particle "tracks". This manual method provides accurate measurements of all motions without a bias towards short or long displacements. The resulting tracking data, which is compiled of coordinate positions for each particle, can then be analyzed to extract particle motion data, such as instantaneous particle velocities, travel times, and travel distances.
