Building on the AGVS dataset series from the previous chapter, our team conducted initial research on computer vision algorithms for airport ground video surveillance. We aimed to leverage unique prior information from airport environments to enhance algorithm performance. This chapter outlines the design principles for airport-specific computer vision algorithms and presents our findings in three key areas: segmentation, recognition, and tracking in airport ground video surveillance.