In　this　article,　an　robust　adaptive　stiffness　visual　servoing　control　method　is　proposed　to　guarantee　a　compliant　physical　interaction　with　a　prescribed　performance　of　the　unmanned　aerial　manipulator　(UAM)　in　the　presence　of　disturbances　and　modeling　uncertainties.　First,　to　address　the　slow　response　speed　of　visual　servoing　control　when　image　features　change,　a　visual　servo　position　controller　with　prescribed　performance　is　adopted　to　enhance　the　convergence　speed　through　the　gradually　convergent　error　boundary.　Second,　considering　the　influence　of　disturbances　and　modeling　uncertainties　caused　by　the　flight　platform　on　the　admittance　accuracy　during　physical　interaction,　an　robust　adaptive　stiffness　physical　interaction　controller　is　introduced,　generating　stable　admittance　command　image　features　to　ensure　a　reliable　interaction　force.　Finally,　physical　comparative　and　real-world　experimental　studies　are　conducted　to　validate　the　feasibility　and　performance　of　the　proposed　method.　The　results　demonstrate　that　the　UAM　under　control　is　capable　of　exhibiting　specified　performance　during　contact　operations,　allowing　smooth　adaptation　to　changes　in　contact　forces.　IEEE