Table 4 Steps of the image analysis. More detailed documentation of the applied algorithms is provided along with the source code https://github.com/robna/ImmobileParticles_MatchAndAnalyse

Initial

particle

detection

• particles get separated from background by thresholding with hysteresis [43]

 – a lower threshold is calculated using Otsu’s method [44], which selects full particle areas

 – areas are accepted as particles if > 10% of their pixels also exceed a higher threshold fixed at 75% maximum brightness

Finding corresponding particles

• image registration: pre- and post-images get aligned

 – particles found in pre- image are matched to equivalent particles in post-image if present

Obtaining particle properties

• matched particles get measured on pre- / post-images

• measurements are used to compute change rates of properties