The invention discloses a wireless communication network intelligent-antenna-covered scene automatic classification and recognition method. The method includes the steps of extracting the length characteristic, the width characteristic, the density characteristic, the collineation characteristic and the shadow characteristic of a community covered by intelligent antennas according to three-dimensional geographic information, conducting excavation analysis on the scene through a clustering tool, selecting certain characteristic parameter values to define and classify the scene so as to obtain the classification of the certain scene covered by the intelligent antennas, and conducting classification recognition on the basis of the classification. According to the method, for the randomness of the existing scene classification standard and the problems of the fuzzy classification boundary and the like, through the combination with the geographic position and the building information of the GIS, the subjective index parameters are defined, the community scene is classified and recognized, and therefore the objectivity and the accuracy which are hard to achieve through manual classification are achieved. Parameters of the intelligent antennas are set according to the scene with the determined classification, and therefore performance of the intelligent antennas can be better given into play, and the requirements of the area for coverage, interference and throughput are met.