In　this　paper,　a　SEIR　epidemic　model　related　to　media　coverage　and　exogenous　reinfections　is　established　to　explore　the　transmission　dynamics　of　COVID-19.　The　basic　reproduction　number　is　calculated　using　the　next　generation　matrix　method.　First,　the　existence　of　equilibrium　points　is　investigated,　and　different　kinds　of　equilibrium　points　indicate　that　the　disease　may　disappear,　or　exist　that　result　in　different　quantity　of　susceptible　individuals,　pre-symptomatic　infected　individuals　and　symptomatic　infected　individuals.　The　stability　of　the　equilibria　is　discussed　by　a　geometric　approach,　and　it　is　found　that　controlling　reproduction　number　to　be　lower　than　1　is　not　sufficient　for　eradication　of　COVID-19.　Second,　transcritical　bifurcation　is　explored,　and　it　is　found　that　improving　the　ratio　of　exogenous　reinfection　may　lead　to　backward　bifurcation　under　poor　medical　conditions,　which　indicates　that　two　endemic　equilibrium　points　appear.　Third,　to　investigate　the　influence　of　parameters　on　the　basic　reproduction,　sensitivity　analysis　is　done　to　choose　relatively　sensitive　parameters,　and　the　parameters　for　treatment　and　media　coverage　are　selected.　An　optimal　control　model　is　established　to　balance　the　treatment　and　media　awareness.　By　exploring　the　existence　and　the　uniqueness　of　the　optimal　control　solution,　the　optimal　control　strategies　are　given.　Finally,　we　run　numerical　simulations　to　verify　the　theoretical　analysis　on　actual　data　of　China,　and　the　data　from　the　four　different　states　of　India　is　used　for　forecasting　the　situation　of　infected　individuals　in　a　short　period.　It　is　found　by　the　simulation　that　the　co-function　of　treatment　and　media　coverage　results　in　the　reduced　number　of　infectious　individuals.