Human-in-loop hybrid enhanced target identification method for unmanned aerial vehicle patrol

The invention discloses a human-in-loop hybrid enhancement target identification method for unmanned aerial vehicle patrol, and belongs to the field of unmanned aerial vehicle patrol. The method comprises the following steps: processing a load image by using a lightweight neural network and a saliency target detection function, and generating an image slice of a region in which a target is detected; the expert evaluation decision module performs personalized display on the slices, and an expert judges a displayed target according to own experience, selects a target which needs to be focused in the patrol process, and generates a control instruction; besides, the network increment upgrading module generates a standard image of the space-time where the target is located according to the selected target, marks the image, adds the marked sample into an original sample library for training, and updates the trained model into the lightweight neural network. According to the invention, the method achieves the quick,