In order to solve the problem that the unmanned aerial vehicle (UAV) is difficult to realize automatic landing at home and abroad, the vision-based navigation technology in the UAV landing process was studied. The landing and landing phases of aircraft was introduced. The image processing technology and pose estimation problem of UAV landing process was analyzed respectively, especially the image feature extraction technology and iterative algorithm of position and attitude estimation during landing. The research results showed that: combining the vision-based navigation with other navigation modes, developing the integrated navigation pose estimation and fusion algorithm with good robustness, high accuracy and good real-time would be the development focus of future vision navigation.