Tunnel　LED　lighting　system　is　an　important　infrastructure　to　ensure　the　safety　and　comfort　of　driving　in　highway　tunnels.　At　present,　traditional　method　for　tunnel　lighting　patrol　of　missing-lamp　(or　faulty-lamp)　condition　is　carried　out　with　manual　labor,　which　has　high　labor　cost　and　low　efficiency.　In　this　paper,　an　automatic　robust　detection　of　missing　LED　lamps　in　highway　tunnels　based　on　vehicle-mounted　video　processing　is　proposed.　This　method　includes　the　following　three　aspects.　Firstly,　a　three-phase　algorithm　was　proposed　to　detect　lighting　LEDs　by　analyzing　the　geometric　structure　of　the　tunnel　interior　environment,　and　the　accurate　detection　of　LED　lamps　in　complex　scenes　with　highlight-interference　in　the　tunnel　was　ensured.　Secondly,　two　kinds　of　quantitative　formulas　are　derived　according　to　the　invariance　constraint　condition　of　line　cross　ratio,　including　the　formula　between　the　two　intervals　of　every　three　adjacent　LED　lamps　and　the　invariance　of　the　Area-Distance?　Ratio　computed　for　different　LED　lamps.　Thirdly,　two　typical　missing　types　for　the　LED　lamps,　including　the　Type　I　of　complete　missing　and　the　Type　II　of　LED　lamps　in　incomplete　illumination　condition,　can　be　detected　axiomatically　through　the　analysis　of　these　two　kinds　of　formulas　cleverly.　The　experimental　results　show　that　this　method　can　robustly　detect　the　missing　LED　lamps　in　the　tunnel　under　complex　conditions,　and　the　accuracy　rate　is　more　than　91　%.　©　2022　Technical　Committee　on　Control　Theory,　Chinese　Association　of　Automation.