Traditional　oral　radiography　is　prone　to　problems　such　as　image　deviation　caused　by　unstable　instruments　held　by　patients　and　inconsistent　projection　positions　by　doctors.　In　order　to　achieve　high　quality　and　stability　of　operation,　a　flexible　intelligent　medical　robot　control　system　has　been　designed.　The　radiation　tube　clamped　on　the　robot　can　track　the　center　position　according　to　the　camera＇s　vision,　completing　image　diagnosis　and　treatment　tasks.　To　this　end,　an　improved　sliding　mode　tracking　control　method　is　proposed,　which　makes　the　robot　system　have　good　invariance　and　anti-interference　performance.　Establish　a　dynamic　model　based　on　the　center　position　of　camera　vision.　A　mapping　model　between　the　control　rate　function　and　joint　torque　was　established　by　combining　the　control　algorithm　with　the　hyperbolic　tangent　function.　The　traditional　control　algorithm　has　been　optimized　to　allow　the　controlled　system　to　move　on　the　predetermined　end　trajectory　of　human　tissue.　Theoretical　analysis　shows　that　the　control　system　converges　asymptotically.　The　experimental　results　validate　the　effectiveness　of　the　control　strategy.　It　improves　the　efficiency　of　intelligent　medical　operations.　Reduce　the　impact　of　subjective　experiential　factors.　This　in　turn　reduces　the　labor　intensity　of　doctors.　©　The　Author(s),　under　exclusive　license　to　Springer　Nature　Singapore　Pte　Ltd.　2024.