The　unwanted　information　such　as　noises　and　interferences　usually　affects　the　quality　of　X-ray　image.　Therefore,　information　extraction　must　be　performed.　The　unsharp　masking　method　is　a　well-known　approach　for　image　enhancement　but　is　very　sensitive　to　noises.　Low-pass　filters　are　usually　employed　to　suppress　the　noise,　but　often　lead　to　the　blurred　edges.　To　overcome　these　problems,　a　sharpening　algorithm　for　X-ray　images　based　on　nonlinear　module　is　proposed.　First,　we　suppress　noise　of　X-ray　images　by　image　averaging;　then　a　lookup　table,　which　is　used　to　sharpen　X-ray　images,　is　set　by　the　grey　difference　between　the　centre　pixel　and　each　neighbouring　pixel.　The　new　method　can　sharpen　images　and　simultaneously　can　have　good　performance　in　noise　reduction.　The　experimental　results　have　demonstrated　the　effectiveness　of　the　proposed　method.