Abstract: There often are a lot of broken contour lines in contour lines images extracted from color topographic maps, which affects vector contour lines production. Hence, research on broken contour line reconstruction has been a active area for a long time. By analyzing the main advantages and disadvantages of the latest methods, and simulating the reconstruction process using artificial intelligence, a novel reconstruction method for broken contour lines based on a compound iterative strategy is proposed in this paper. In this method, the inner regularity of broken lines' caps and shapes is summarized, then an incremental, circular and iterative strategy for end-point matching and connecting is applied, in which all the end-points are classified into different intervals and shape classes. The key algorithms for obtaining end-point directional angles and directional differences were significantly improved. In experiments, all end-point reconstructions were complemented, using this method. These experiments indicate that this method holds better stability and applicability, and acquires excellent effects as measured by quantitative indices.