Abstract: Accurate mapping algorithm is the core of ArUco landmark array localization. Most of the existing mapping algorithms based on offline processing or manual calibration behave inefficiently and time-consuming. In addition, the existing online mapping algorithms can though guarantee real-time performance, but show low accuracy, large cumulative error and sensitive to false detection. In order to overcome those shortcomings, a monocular ArUco-SLAM algorithm was proposed based on parallel mapping and synchronization optimization in this paper. Firstly, the closure principle of coordinate transformation was used to build and correct the global map online. And then, a linearity optimization algorithm was used to minimize the projection error in the corners and to achieve global optimization for key frame of the map. A synchronization optimization algorithm was used to match the online map and optimized map, and to reduce the cumulative error of the global map. Finally, several contrast experiments with other mapping systems were carried out. The experimental results show that the proposed system can not only establish high-precision map, but also realize online mapping and real-time camera localization.