Skip to Main Content
Reversible watermarking is suitable for hiding data in 2-D vector maps because the distortions induced by data embedding can be removed after extracting the hidden bits. In this paper, two reversible data-hiding schemes are explored based on the idea of difference expansion. The first scheme takes the coordinates of vertices as the cover data and hides data by modifying the differences between the adjacent coordinates. The scheme achieves high capacity in the maps with highly correlated coordinates. Instead of the raw coordinates, the second scheme adopts the manhattan distances between neighbor vertices as the cover data. A set of invertible integer mappings is defined to extract manhattan distances from coordinates and the hidden data are embedded by modifying the differences between the adjacent distances. For those maps where distances exhibit high correlation, this scheme shows better performance than the former one, both in capacity and invisibility. Three different maps with distinct features are used for the experiments. The results indicate that two proposed schemes suit different types of maps, respectively, according to the correlation of the selected cover data. Both schemes are strictly reversible. In addition, they can be slightly robust for low amplitude distortions by selecting higher digits for data hiding. The potential applications of proposed schemes may include map data authentication, secret communication, etc.