Reversible Watermarking Algorithm for Vector Maps Using the Difference Expansion Method of a Composite Integer Transform

Current reversible watermarking algorithms for 2D vector maps do not consider the shape characteristic of 2D vector maps.The obvious disadvantage of these techniques is that vector map quality may be severely degraded with low data hiding capacity.This paper presents an effective reversible watermarking algorithm for 2D vector maps.In the algorithm,through analyzing the characteristic of vector maps,multidimension vectors are constructed from the subsection monotonicity of coordinates of curves and polygons,and a composite difference expansion integer transform is proposed and used for the multidimension vectors.Thus,a large payload can be embedded into a vector map with low distortions in the monotonous transition zones of a stego-map.In addition,by setting the threshold of error tolerance,the length of embedded data is controlled.Theoretical analysis and experimental results show that our algorithm allows for hiding high payload with low distortions.The potential applications of proposed scheme may include vector map authentication,secret communication,etc.