The invention discloses a method and a system for detecting an image target of an unmanned aerial vehicle. The method comprises the following steps: judging whether the difference value between the reference frame of the target and the current frame exceeds a threshold value, wherein, the reference frame is an adjacent first frame of the current frame; extracting respective features of the reference frame and the current frame if the difference value exceeds a threshold value; transferring characteristics of the reference frame to the current frame through the optical flow network model; combining the current frame and the features transferred from the reference frame according to different preset weights, wherein, the weights in the feature channels are fixed spatial weights; detecting the enhanced features to get the results of target detection and recognition and semantic segmentation results. The invention solves the technical problem that the processing precision and the efficiency can not be taken into account when detecting and identifying the target of the unmanned aerial vehicle image in the prior art.