A method based on analysis of the region of movement and the functioning of the acto-myosin cytoskeleton has been elaborated to quantify and classify patterns of organelle movement in tobacco pollen tubes. The trajectory was dilated to the region of movement, which was then reduced to give a one-pixel-wide skeleton, represented by a graph structure. The longest line in this skeleton was hypothesized to represent the basic track of the organelle along a single actin filament. Quantitative features were derived from the graph structure, direction of movement on the longest skeletal line, and distance between skeletal line and particle. These features corresponded to biological events like the amount of linear movement or the probability of attachment of an organelle to the actin filament. From 81 analyzed organelle trajectories, 17 had completely linear, 17 had completely non-linear, and 47 had alternating linear and non-linear movement. Selected features were employed for classification and ranking of the movement patterns of a representative sample of the population of organelles moving in the cell tip. The presented methods can be applied to any field where analysis and classification of particle motion are intended.