Abstract: A multiple watermarking algorithm of vector map is proposed to achieve the copyright protection with the high-precision requirement. This algorithm uses two types of watermarks: zero-watermark and reversible watermark. First, the vertices are clustered according to the mesh density and the attribute entropy. The density center points are chosen to embed the zero-watermark bits, and the feature points within a relative distance to the center are applied to identify regions for the reversible watermark insertion. Then, the copyright watermark image is scrambled with Arnold transformation to improve security. And in the zero-watermarking scheme, the improved zero-bit dynamic extension algorithm is used to modify only the 〖WTBX〗x〖WTBZ〗 (or 〖WTBX〗y〖WTBZ〗) coordinate values of the center points, which will need less points to embed the copyright information and also can improve robustness against precision reduction. In the reversible watermarking scheme, the improved difference expansion and shifting algorithm is used. Different reversible watermarking formulas are applied to different interval points to embed a high capacity reversible watermark. Only the 〖WTBX〗x〖WTBZ〗 (or 〖WTBX〗y〖WTBZ〗) coordinate values of the feature points are hided copyright information to reduce the effect of the map on quality. The experimental results show that the proposed multiple watermarking algorithm can better solve the contradiction between vector map accuracy and watermark robustness.The multiple watermarking scheme based on zero-watermark and reversible watermark has better attack resistance than single watermarking scheme,which is eligible for copyright protection of vector map with high data precision requirement.