Image Processing and Photographic Mosaic

《Image Processing and Photographic Mosaic》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、Lab4:ImageProcessingandPhotographicMosaicEE299Winter2008Due:Exercises1-2aredueonorbefore8February,andtheremainderofthelabisdueonorbefore15February2008.ObjectiveThepurposeofthislabistolearnaboutdigitalimagesandperformsomebasicimageprocessinginMATLAB.YouhavealreadydonesomeworkwithimagesinMATLABi

2、nLab1.Inthislab,youwilllearnafewmoretoolsforhandlingimagesanddevelopabetterunderstandingofthetoolsthatyouhaveused.Morespecifically,thislabcovers:1:representinganddisplayingimagesinMATLAB;2:simpletechniquesforimagefilteringandtransformation;3:howtocreatesyntheticcolorimages;and4:creatingaphotogra

3、phicmosaic(animagecreatedfrommanymuchsmallerimages).Therearefourexercisestotalinthislab:exercises1-3allowyoutopracticeloading,displaying,andchangingimagesinMATLAB,andinexercise4youwillmakeyourphoto-mosaic.Prelab:Toensurethatyoucancompleteeverythingwithintheallottedtime,beforeyourlabsectionyous

4、hould:•Readthelabbeforecomingtolab.•Trytofinishparts1-2(ideally1-3)duringthefirstweekoflab.•Beforethesecondweekoflab,findadigitalpicturethatyouwouldliketoturnintoaphoto-mosaic,andmakesurethatitisavailabletodownloadtoalabcomputer.Ideally,youshouldpicksomethingthatisfairlyhighresolutiontostartwith.

5、Teammatescanhavethesamepictureordifferentones.Optionalformoreambitiousstudents:Ifyouwanttousetheirownimagecollectiontobuildthemosaic,youshouldhavethatcollectedinonedirectorythatyoucanaccessfromtheSieg232lab,orburnaCDaheadoftime.1RepresentingImagesinMATLABInthissectionwewilltalkabouthowtoread,di

6、splayandwriteimagesinMATLAB,aswellashowtomanipulatesubimages.Asapreliminary,itisimportanttounderstandthatadigitalimage(colororgrayscale)isjustabunchofnumbers,butthenumbersthemselvescanbecharacterizedusingdifferentformats.Digitalsystemstypicallyrepresenttwotypesofnumbers:integerandfloatingpoint.T

7、herearemanytypesofeachinMATLAB,buttheoneswewillfocusonaredouble(floatingpoint)anduint8(8bit1Table1:Differencesbetweendoubleanduint8inMATLABdoubleuint8typefloatingpointintegerbits648rangeroughly-1.8e+308to1.8e+3080to255unsignedinteger).Most