For　the　problem　that　industrial　robot　have　a　poor　ability　to　perceive　external　environment,　and　not　able　to　change　the　way　of　operation　and　path　of　transformation　timely　as　working　condition　changes.　This　paper　selects　color　sensor　as　the　medium　of　recognition,　introduces　the　embedded　system　Raspberry　Pi　to　6　axis　industrial　robot　grabbing　assignments,　and　puts　forward　a　six　axis　industrial　robot　intelligent　recognition　model　with　color　perception,　dynamic　decision-making　and　other　multi-function.　In　the　model,　the　color　sensor　signal　will　be　resolved　by　color　processing　algorithm　in　embedded　system,　and　consequently　color　information　of　material　on　the　conveyor　can　be　obtained.　The　color　information　after　processing　by　the　embedded　system　will　be　converted　to　digital　signals.　Then　input　the　digital　signal　into　robot　controller.　Finally,　robot　controller　chooses　corresponding　motion　path　according　to　the　input　signal.　The　experimental　results　show　that　the　six　axis　industrial　robot　intelligent　recognition　system　can　distinguish　colors　of　material　with　almost　100%　accuracy,　and　adjust　operation　way　and　path　in　0.72　s　time　rapidly.　Efficiency　could　be　improved　by　more　than　75%.　©　2017　IEEE.