In response to the challenges of real-time efficiency and safety in parafoil homing trajectory planning for large-scale airdrop operations, we introduce an adaptive trajectory optimization technique for pilot-type parafoils with terrain analysis. To begin with, a terrain analysis algorithm is utilized to select a nominal landing point with undulating terrain suitable for the airdrop mission. Feature identification was then utilized for conducting the parafoil homing trajectory planning for the pilot-type parafoil. The parafoil trajectory planning problem is constructed parametrically taking into account of the working characteristics of the pilot-type parafoil and mission requirements. The trajectory optimization is then carried out using the interior point method coupled with the obstacle penalty function approach. Real-time image recognition is performed using map data collected by the pilot-type parafoil to determine the optimal landing point for subsequent parafoils. This methodology offers an alternative means of self-detecting the suitable landing point and self-planning the reliable trajectory. It provides a solution that guarantees both the accuracy and real-time performance enhancement of the parafoil cluster, which improves the reliability and safety of the homing task.