In　this　study,　a　new　robust　homography-based　visual　tracking　control　approach　for　the　quadrotor　unmanned　aerial　vehicle　(UAV)　is　developed.　Specifically,　employing　the　homography　matrix　as　feedback,　a　hierarchical　homography-based　visual　servoing　(HBVS)　scheme　with　a　new　command　attitude　extraction　method　to　account　for　the　underactuation　of　UAV　is　proposed.　On　this　basis,　a　smooth　hyperbolic　tangent　function　is　fulfilled　as　an　augmented　part　of　the　backstepping　control　scheme,　which　guarantees　the　non-negative　total　thrust　and　avoid　singularity.　Additionally,　a　cascaded　filter-based　estimator　and　adaptive　laws　with　integrable　functions　are　embedded　to　counteract　uncertainties　including　external　perturbations,　unknown　acceleration　of　the　moving　target,　and　unknown　image　depth,　and　to　facilitate　the　system＇s　asymptotic　stability　simultaneously.　The　theoretical　analysis　testifies　that　the　whole　close-loop　system　is　asymptotically　stable.　Simulations　further　verify　that　the　proposed　HBVS　controller　can　realize　the　visual　tracking　with　a　superior　performance.　©　2022　The　Franklin　Institute