home *** CD-ROM | disk | FTP | other *** search
/ Atari Mega Archive 1 / Atari Mega Archive - Volume 1.iso / graphics / aim_xtra.arc / AIMTUTOR.DOC < prev    next >
Text File  |  1988-04-07  |  46KB  |  957 lines
  1. 06600020202800
  2. 1AIM TUTORIALPage # ofFile: AIMTUTOR.DOC
  3. 2
  4. F0110030
  5. 9[...................................................]011
  6. ü
  7.  
  8.  
  9.  
  10.  
  11.  
  12.  
  13.  
  14.  
  15.  
  16.  
  17. ëAIMTUTORIAL
  18. 9[...................................................]011
  19. Ç
  20.  
  21. by
  22.  
  23.  
  24. TomHoeksmaandAdHerweijer
  25.  
  26.  
  27.  
  28. DelftUniversityofTechnology
  29. FacultyofAppliedPhysics
  30. PatternRecognitionGroup
  31. Lorentzweg1
  32. 2628CJDELFT
  33. theNetherlands
  34. ëIntroductionê
  35. Ç
  36. TheimageprocessingsoftwarepackageAIM(AtariImageManager)
  37. hasbeendevelopedatDelftUniversityofTechnologybyFrans
  38. Groen(the'imageprocessor')andwasadaptedfortheAtariST
  39. computerincooperationwithRobertdeVries(theGEMprogram⑨
  40. mer).BothweremembersoftheSignal-/System-techniquesGroupin
  41. theFacultyofAppliedPhysics,Fransasaseniorstaffmember
  42. andRobertasagraduatestudent.Recently,Franschangedjobsto
  43. becomeprofessorofAppliedInformaticsintheUniversityof
  44. Amsterdam.
  45.  
  46. AIMisdesignedtobeanentrylevel'appetizer'fortheartof
  47. imageprocessing.Itgivestheinterestedlaypersonan
  48. opportunitytoacquainthimselfwiththetechniquesand
  49. terminologyusedinthisdynamicfield.Theexperiments
  50. describedmayberegardedasalowerboundofwhatcanbe
  51. accomplishedwithAIM,theupperboundbeinglimitedbythe
  52. user'simaginationonly.
  53.  
  54. Thistutorialisnotafullcourseondigitalimageprocessing.
  55. SuchcoursesarefrequentlyorganizedbythePatternRecognition
  56. Group.Forcourseinformation,pleasecontacttheSecretaryof
  57. thePatternRecognitionGroup,FacultyofAppliedPhysics,
  58. Lorentzweg1,2628CJDelft,theNetherlands,Phone015-781416.
  59.  
  60. Notethattheoperationsusedintheexamplesanddemoscovera
  61. minorpartofAIM'scapabilitiesonly.Acomprehensivedescrip⑨
  62. tionofalltheoperationsavailableisgiveninafilecalled
  63. AIMANUAL.DOC.Likethistutorialitiscontainedinacompressed
  64. file(DOCS_IMS.ARC),whichcanbefoundinthefolderCOMPRESSon
  65. thedistributiondisk(togetherwithadecompressingprogram).
  66.  
  67. Imageprocessingisnotparticularlydifficult;manypeoplewill
  68. havehadtheirfirstimageprocessingexperienceinthe
  69. photographicdarkroom.Note,however,thatAIM,beingdedicated
  70. todigitalimageprocessing,requiressomebasicknowledgeof
  71. digitalsystems.Sothoseofyouwhountilnowhavesuccessfully
  72. avoidedlearningaboutbits,bytesandbinarynumbers,be
  73. preparedtofacetheunavoidable!The15minutesorsospenton
  74. readingabouttheseconceptswillberichlyrewardedbyabetter
  75. understandingofAIM'soperations,ifnotbythedeepadmiration
  76. ofyourenvironmentforthenewlyacquiredvocabulary.
  77.  
  78. AIMisaresultofmanyyearsofresearchanddevelopmentinthe
  79. PatternRecognitionGroup.Manyoftheimageprocessingroutines
  80. werecontributedby(a.o.)GertvanAntwerpen,FransGroen,Piet
  81. Verbeek,BenVerwer,LucasvanVlietandIgorWeber.This
  82. introductionwaswritten(ifnotcompiledfromexistingtexts)by
  83. TomHoeksmaandAdHerweijer.
  84.  
  85. PreviousversionsofAIMhavealreadyextensivelypenetratedthe
  86. AtariSTcommunity.However,asthedocumentationsuppliedwith
  87. thosefirstpublicdomainversionswasratherlimited,the
  88. explorationofthesoftwarehasbeenamere'imageprocessing
  89. adventure'tomostofthe'earlybirds'.Theauthorshopethat
  90. thisintroduction,togetherwiththerelatedexamplesanddemos
  91. includedonthedisk,willenableallprospectiveusers(early
  92. birdsandnewcomersalike)tobetterunderstandwhatwas,resp.
  93. isgoingon.
  94. Ç
  95. Feedbackplaysanimportantroleintheeffectiveuseofimage
  96. processingaswellasinsoftwaredevelopment.Therefore,we
  97. inviteyoutosendyoureventualcommentsandsuggestionsto:
  98.  
  99. Prof.dr.ir.F.C.A.Groen
  100. c/oDelftUniversityofTechnology
  101. FacultyofAppliedPhysics
  102. PatternRecognitionGroup
  103. Lorentzweg1,2628CJDELFT
  104. theNetherlands
  105.  
  106. ë
  107.  
  108. ëWHATCANWEEXPECTOFIMAGEPROCESSING?Ç
  109. 9[...................................................]011
  110.  
  111. MaybeyouhavealreadyhadsomeexperiencewithAIMorseen
  112. somebodyelseworkwithit.Youmayhavewondered,then,how
  113. imageprocessingcouldmakefeaturesappearinthedisplaythat
  114. weren'ttherebeforetheoperation.Textthatatfirstwas
  115. unreadable,suddenlycouldbereadeasily;frecklesinagirl's
  116. face,thatdidn'tshowintheoriginaldisplay,could(toher
  117. distress!)becountedaftersomeappropriatekeystrokes.Youwere
  118. perhapsconvincedtohaveacquiredalittlemagician,hiddenin
  119. thattinyfloppydisk.
  120.  
  121. Bewarned,however!Imageprocessingcanneverreproducefeatures
  122. oftheoriginalimage,thatsomehowhavebeenlost.So,be
  123. preparedforsurprises,butdon'texpectmiracles!Image
  124. processingcanenhanceparticulardetailsinanimage,but
  125. usuallyotherdetailswillhavetobesacrificed.
  126.  
  127. If'Apictureisworthathousandwords',imageprocessingallows
  128. youtoarrangethosewordstomakethepicturerevealits
  129. (secret)'stories'.Butremember:onepicture'sthousandwords
  130. aredifferentfromanother's.Everystorywillinheritthe
  131. characteristicsofthepictureitwasextractedfrom.Theartof
  132. imageprocessingis:tomakethepicturetelltheonestory
  133. that'susefultoyou.
  134.  
  135.  
  136. ë
  137. STAGESINANIMAGEPROCESSINGAPPLICATIONÇ
  138.  
  139. Inalmosteveryimageprocessingapplicationwecanrecognize
  140. somestages,thateventuallyleadtothefinalresult:ürecordingÇ,
  141. üpreprocessingÇ,üsegmentationÇ,üpostprocessingÇ,üanalysisÇandüinter⑨Ç
  142. üpretationÇ.
  143.  
  144. Probablythemostimportantstageistherecordingofthe
  145. originalpicture.Asstatedabove,informationthatgetslostin
  146. therecordingprocesscanneverberecovered.Fortherecordinga
  147. video-cameraisrequired,capableofconvertingthelight,
  148. reflectedfromtheobject,intoanelectricalsignal.Asthe
  149. videosignalcannotbereaddirectlybythecomputer,thissignal
  150. mustbesampledanddigitizedandthedigitalinformationmustbe
  151. transferredinanorderlymannertothecomputer'smemory.This
  152. isdonebyaspecialpieceofhardware,aso-called'üframe
  153. grabberÇ'.VariousframegrabbersareavailablefortheAtariST.
  154. InthePatternRecognitionGroupmonochromeandcolorframe
  155. grabbershavebeendevelopedthateventuallywillbemade
  156. commerciallyavailable.
  157.  
  158. Aftertheimageisstoredinthecomputeritmustbepre⑨
  159. processed.Heretheusertriestoenhanceorfilterout
  160. particularwantedresp.unwantedfeatures,tosatisfy
  161. requirementsoffurtheranalysis.Examplesofpreprocessingare
  162. thecorrectionforchanginggreyvaluesinthebackground
  163. (shading),suppressingnoise,determinationofedgesinthe
  164. image,etc.
  165.  
  166. ÇInordertobeabletodeterminepropertiesofsingleobjectsin
  167. theimage,theseobjectsmustfirstbemadedetectable.Ifagrey
  168. valueimagecontainsdarkobjectsonalightbackground,
  169. segmentationcan,forinstance,resultinabinaryimagewith
  170. whiteobjectsinablackbackground.
  171.  
  172. Postprocessingcanimprovetheresultofthesegmentation.
  173. Examplesarethemorfological(binary)operations.Holesin
  174. objectscanbeclosed,edgespeeledoff,etc.
  175.  
  176. Intheanalysisstage,measurementscanbeperformedonthe
  177. segmentedobjects.Examples:area,perimeter,curvature.
  178.  
  179. Intheinterpretationstagetheresultsofthemeasurementsare
  180. evaluated.
  181.  
  182. An'intelligentrobot'inafactor