Wind-blown sand flow is the basic phenomena that have profound influences on the environment, and the experimental study on the velocity distribution of the sand particles is very important for the understanding of this phenomena. Among the various experimental techniques, particle tracking velocimetry (PTV for short) is one that attracts more and more attentions.

In this paper, an algorithm based on the Self-Organizing Maps network is established for PTV in a wind-blown sand flow, and the processing results by the algorithm prove its ability to capture the characteristics of the concerning flow field.