In　metallic　plate　structures,　a　lamination　defect　is　a　type　of　internal　defect　that　is　commonly　generated　in　the　manufacturing　process,　which　is　mostly　parallel　to　the　surface　of　the　plate.　This　study　presents　a　technique　of　imaging　a　lamination　defect　in　a　metallic　plate　by　using　Lamb　waves　based　on　the　local　wavenumber　domain　analysis　method.　Owing　to　the　processing　technique,　an　aluminum　beam　structure　is　utilized　to　represent　the　metallic　plate　for　the　detection　of　the　lamination　defect.　A　three-dimensional　finite-element　model　of　an　aluminum　beam　with　a　rectangular　lamination　defect　is　constructed　with　MatLab　and　ABAQUS　commercial　software.　The　propagation　characteristics　of　Lamb　waves　are　discussed　by　analyzing　the　simulated　wavefield.　A　complicated　wave-propagation　phenomenon　of　multiple　reflection　and　transmission　is　observed　between　the　interaction　of　Lamb　waves　and　the　lamination.　By　performing　local　wavenumber　domain　analysis　on　the　wavefield,　the　wavenumber　distribution　is　obtained,　from　which　the　location　and　profile　of　the　lamination　defect　are　identified.　An　experimental　study　is　carried　out　to　verify　the　numerical　result.　A　fully　non-contact　detection　system　consisting　of　an　electromagnetic　acoustic　transducer　and　laser　ultrasonic　inspection　system　is　established　to　inspect　the　artificial　machined　beam　structure.　The　experimental　result　is　in　good　agreement　with　the　numerical　result.