Abstract: Anewevaluationfunctiontocomputeremotesensingimagedefinitionwasproposedaccordingtoengineeringapplicationsandfeaturesofremotesensingimage,toovercometheshortagesoftraditionalimagedefinitionevaluationfunctions.Forlargedataanddifferentscenerydensityanddistributionofremotesensingimages,thedefinitionevaluatingprocesswasseparatedastargetareadetectionandcharacteristicparameterextraction.Firstly,detectionfunctiondetectedwholeimagetogettherichsceneryandobviousedgeareaastargetarea.Thenagraygradientfunctionmadeimagegraypretreatmentandextracteddefinitioncharacteristicparameterstoevaluateimagedefinition.Threedifferentkindsofremotesensingimages,namely,emulateddefocusingimage,emulatednoiseimageandonorbitsatelliteimagewereusedintheproposedmethod,toverifyitseffectivenessandaccuratenessbycomparingitwithfourtraditionalmethods.Theanalysisandemulationresultsshowthattheproposedmethodcanmeetthefourbasicdemandstoevaluatingremotesensingimagedefinitionandtacklethedeficienciesoftraditionalfunctionsinevaluatingdifferentremotesensingimages,itisavailabletoevaluateremotesensingimagedefinition.

Key words: imagedefinition, targetarea, characteristicparameter, pretreatment, imageevaluation, spaceremotesensing