Abstract: Some complex problems have to be faced during medium-small scale map making from large scale geological map, such as low efficiency to merge massive geological polygons, complicated processing procedures, and being difficult in assuring data quality.This paper proposes a novel intelligent generalization method for geological body based on cartographic and artificial intelligence.Dissolving, spatial aggregation, customized fusion, elimination and exaggeration are employed to perform intelligent processing integrating map generalization and intelligent algorithms.iMapower(MapGIS version)implements the above generalization methods in corresponding modules, which can run automatically or in the form of human-computer interaction.It is very easy to complete most of complete geological body mapping tasks just clicking one time.Using real data to verify the proposed method is effective and can be corrected during mapping from large-scale map to small-scale map, and it can save at least 85 percent of geologic body generalization workload.It can greatly improve the work efficiency and assure the data quality.Furthermore, this way can play an important role in pursuing intelligent compilation of geological map.