The invention discloses an unmanned aerial vehicle flight landing system and method based on image processing. The system comprises an unmanned aerial vehicle and an unmanned aerial vehicle take-off and landing station. The unmanned aerial vehicle take-off and landing station comprises a carrying table arranged on the ground. A precise positioning and fixing device assisting in landing of the unmanned aerial vehicle is mounted on the carrying table. A holder is mounted on the precise positioning and fixing device assisting in landing of the unmanned aerial vehicle. A camera is mounted on the holder, and a camera lens of the camera faces upward. The camera and the holder are both connected with a computer. The computer controls rotation of the holder to drive the camera to rotate, and accordingly the shooting direction of the camera is adjusted. The computer controls the camera to shoot pictures. The computer is in communication connection with the unmanned aerial vehicle through a first wireless data transmission module. The computer is further connected with a display terminal. The unmanned aerial vehicle flight landing system and method have the beneficial effect that an all-weather centimeter-level precise landing function of the multi-rotor unmanned aerial vehicle is achieved based on image processing by adoption of a multistratum PID control technology.