The present invention provides an unmanned aerial vehicle real-time map reconstruction method. By applying an unmanned aerial vehicle, the three-dimensional maps of the surrounding environment of the unmanned aerial vehicle are obtained, and the own positioning of the unmanned aerial vehicle is realized. According to the method, the unmanned aerial vehicle flies in the unknown airspace, and on the condition that the own position of the unmanned aerial vehicle is not clear, the own positioning of the unmanned aerial vehicle is obtained and the three-dimensional maps are constructed by detecting and comparing the environmental information, and extracting and matching the feature information of the environment. The unmanned aerial vehicle real-time map reconstruction method also comprises the geographic information fusion and an image mosaic technology. The unmanned aerial vehicle real-time map reconstruction method of the present invention can be applied to the fields, such as the battlefield demands, for example the unmanned aerial vehicle autonomous flight and attack, the disaster emergency rescue, such as fire hazard, earthquake, flood, etc., the unmanned aerial vehicle security administration and monitoring, etc. Compared with the conventional method, and according to the present invention, a position drifting amount, an angle drifting amount or an absolute error is above the average, and more environmental information is obtained. Moreover, compared with a full-density method, an experimental precision of the method still can reach a usage standard, and the method can be operated on a CPU directly and is good in real-time performance.