Local helioseismology provides an opportunity to look below the surface and see the vertical structure of supergranulation. We present a concept of a (3 + 1)-D segmentation algorithm capable of recognising individual supergranules in a series of helioseismic 3-D flow maps.
As an example, we applied this method to the state-of-the-art data and derived descriptive statistical properties of segmented supergranules, which are in agreement with generally accepted values. We also present preliminary results obtained on the topic of the three-dimensional structure and evolution of supergranulation.