A　novel　road　slope　monitoring　and　early　warning　system　was　developed　by　integrating　3D　image　recognition　technology　and　3D　numerical　modeling　technology　for　monitoring　and　predicting　slope　deformation.　It　was　applied　for　monitoring　and　early　warning　for　a　road　slope　in　Nanping,　Fujian,　China.　The　system　consists　of　equipment　information　management,　data　management,　forecast　and　early　warning,　information　release　and　model　visualization　modules.　It　can　carry　out　point-surface-body　monitoring,　3D　model　visualization,　damage　trend　prediction　and　early　warning　of　landslides.　From　year　2022,　three　rainfall　events　(January　22:19.8 mm/day,　April　30:29.2 mm/day,　and　May　27:31.8 mm/day)　were　predicted　and　verified　using　this　system.　The　results　show　that:　(1)　The　displacement　results　of　the　severely　deformed　region　predicted　by　numerical　simulation　are　similar　to　the　displacement　results　of　image　recognition.　With　the　increase　in　rainfall　intensity,　the　surface　layer　of　some　areas　shed　0.18–1.1 m,　and　the　error　was　within　15%;　(2)　The　predicted　position　of　the　deformation　area　is　consistent　with　the　position　identified　by　the　image,　all　of　which　are　at　the　top　of　the　slope,　and　a　small　part　is　on　the　right　side　of　the　middle　of　the　slope;　(3)　The　fluctuation　range　of　the　displacement　tangent　angle　of　the　three　rainfall　events　is　0–44.32°,　the　slope　is　relatively　stable　as　a　whole,　and　it　is　in　the　stage　of　no　warning.　The　successful　implementation　could　provide　a　reference　for　slope　disaster　monitoring　and　early　warning.　©　2023,　The　Author(s),　under　exclusive　licence　to　Springer-Verlag　GmbH　Germany,　part　of　Springer　Nature.