With　the　continuous　development　of　intelligent　transportation　technologies　such　as　autonomous　driving　and　navigation,　accurate　perception　of　road　markings　becomes　crucial.　However,　due　to　limitations　in　sensor　perspectives　and　obstacles　blocking　the　view,　the　scanned　point　cloud　of　road　markings　is　often　incomplete,　potentially　leading　to　erroneous　decisions　in　intelligent　transportation　systems.　Therefore,　it　becomes　imperative　to　recover　complete　road　markings　from　these　incomplete　point　clouds.　This　paper　proposes　a　text-guided　road　marking　completion　method,　which　integrates　text　information　with　point　cloud　data　using　attention　mechanisms.　By　leveraging　text　information　to　guide　the　network　in　completing　road　marking　point　clouds,　the　proposed　method　aims　to　enhance　the　perception　accuracy　and　completeness　of　road　markings.　Experimental　validation　on　road　marking　datasets　demonstrates　the　effectiveness　and　feasibility　of　the　proposed　approach.　©　2024　IEEE.