Author/Authors :
Wang, Hui School of Information and Control Engineering - China University of Mining and Technology - Xuzhou, China , Li, Shiyin School of Information and Control Engineering - China University of Mining and Technology - Xuzhou, China , Hu, Xiaojuan School of Physics - China University of Mining and Technology - Xuzhou, China , Xu, Hui School of Computer Science and Technology - China University of Mining and Technology - Xuzhou, China , Lu, Zhaolin Advanced Analysis and Computation Centre - China University of Mining and Technology - Xuzhou, China
Abstract :
Scanningelectronmicroscopy(SEM) playsanimportantroleintheintuitiveunderstandingofmicrostructuresbecauseitcanprovideultrahighmagnification. Tensorhundredsofimagesareregularlygeneratedandsavedduringatypicalmicroscopyimagingprocess. Giventhesubjectivityofamicroscopist’sfocusingoperation,blurrinessisanimportantdistortionthatdebasesthequalityofmicrographs. Theselectionofhigh-qualitymicrographsusingsubjectivemethodsisexpensiveandtime-consuming. Thisstudyproposesanewno-referencequalityassessmentmethodforevaluatingtheblurrinessof SEM micrographs.ThehumanvisualsystemismoresensitivetothedistortionsofcartooncomponentsthantothoseofredundanttexturedcomponentsaccordingtotheGestaltperceptionpsychologyandtheentropymaskingproperty. Micrographsareinitiallydecomposedintocartoonandtexturedcomponents.Then,thespectralandspatialsharpnessmapsofthecartooncomponentsareextracted.Onemetriciscalculatedbycombiningthespatialandspectralsharpnessmapsofthecartooncomponents. Theothermetriciscalculatedonthebasisoftheedgeofthemaximumlocalvariationmapofthecartooncomponents. Finally,thetwometricsarecombinedasthefinalmetric.Theobjectivescoresgeneratedusingthismethodexhibithighcorrelationandconsistencywiththesubjectivescores.
Keywords :
No-Reference Quality Assessment Method , Blurriness , SEM Micrographs , Multiple Texture , Scanningelectronmicroscopy(SEM)
