Abstract: In map generalization, it’s important to simplify linear objects. As a patch against other optimization algorithms, ant colony optimization algorithm in solving combinatorial optimization problems has some advantages, such as faster convergence and is more closed to problem-solving habits of the human. The simplification of linear objects can be looked upon as an optimization problem about partial selection. According to the purpose for simplifying linear objects, the standards on map production and the basic principles of ant colony optimization algorithm, the simplification of linear objects is explained as a kind of combinatorial optimization problem, and the constraints that should be satisfied in the simplifying process are described by mathematical formulae at first in this paper. Then a model of automated simplification using ant colony optimization algorithm is put forward in detail and the key steps are given. In order to improve operational efficiency of the algorithm, long-term taboo list and the local search strategy are introduced. Finally, the algorithm is tested and compared with Douglas algorithm, the results demonstrate that the proposed model to the automatic generalization of linear objects is feasible and effective, and the objective function values obtained from ACO algorithm are better within the same geometric tolerance. Using the proposed algorithm the basic geometric characteristics of linear objects are maintained, and a higher compression ratio is achieved while taking into account length deviation and vector deviation.