Department of Computer Science & Engineering Jahangirnagar University ASSIGNMENT Course Name: Digital Image Processing Course Code: CSE-407 Submitted To: Dr. Liton Jude Rozario Professor CSE, JU Name: Anower Hossain Roll: 2118 Exam roll: 170520 4" year 1" semester Session: 2016-17 Date of Submission: 11-06-2021 Scanned with CamScanner4). Whol is an image 4 Mentiond rome. @pplient 10M ot -meslieg| image Prescenhing 4] > Tmage : An image th et hing bud a tury dim 2ntiong| pignal . 44 is debined by the methemati- cal gn (x TeiRontalh Day DY where AY Gre ty co-ordinates Nand Vercheglly , image brightnens Valuer ean be any ned Mumberd in Hy Wenge ef AoC uhite) 9:0 block) Tay An image ib “He mahi Sf gray Ya)teoy AP plicadians ob —mediea\ imoge Proceabing : The digital imoge haa “many Fieldy jn the “medical field Aveh as — 1) Gemma Ray Imoging ") PET Sean Mm) X-Ray imag ing ™) Medical eT Sean ‘ “Y uv maging OS Medicine. Scanned with CamScannerb) Dishiaguish bel ween origh| neh and Crlnayt. Give -methematieal eypressions “Ho enhance Colin brightness and eprinas! = 6P a@h image . Answer! An image han proper cantrant and brigh ‘rebh bore eany Viewing. Brightness reaker to th overall lightnep or dartKneny of an im gga Corhaa} 15 the dibberent brightness between OS yee on Tei ona. : Prightness Contras PBrightners ip the abn > Contras} qualsty wahich only depends én R&® rvaluer of image is tabrhve auelity which depend gn the beele Ground where gbyect placed, > Brightnenp tb Aten @l> di is qvarithy thet exes mbed ving the RGB object Teetognition ran -, Stale . bnightnesd, pBepands on som of tel > depends on =minimum and amplitode { ROS pixels MAXIMUM Pixel intensity Mathematien\ exmmsssien’ The basie. eatent gnl Scanned with CamScanner$0) =u+ 8 for. anion Agpomahion of St moda-Reet eng Sormula > Pig = 1-102) +3945 . Wharee , b = commas brightaenn ©) whet iy Sam ging (es) and quantization? Answers Sen ‘orightnens | @nd ‘contra gling rebers fy tHe digit ain eon} 9 The Sampling teqte will of course abled Tasclutien ok the image . iaveuy Fing\ Quant Ration reben to Hu grayscale God tp TEP te sent “the image The transition behoeon CaorHinuays vaheos 64 image £0 and its digital eqvivalest ig) called quantization . From #a-o4, Sampling ant quantyect on: 4b! We “inenedse | the Tesolvtign ot the image fore example, 320%4e0 the VmOaR wil) be more THCogNi gy ble Ab we decrease TOE) o+ 6m be les, mecranionhle it wilt Scanned with CamScannerG2 What id meant by Histogram oh On img oP he histogram ib Wed fore Arta phieg! etion of on image . A: Qroph is plet by ben of Pixel» bon each Valuan . Newadar histogram 18 presen: tn digital Comoran axis TeD Tee sen} ~fong) Vertiotong Uhsray y Wied — repre send MWimben Ef Pixels . Hew 9 histegram carttasted 4s > Contrast Te", the tinst atep ig fo bin the Tange thet in divide the en} ire Tenge ob velues aartieh of imbenvals. “The Niohg tam Cen be cambrueted ©Y pimp) Te number of timen LGA Tey -rtale Valine | Seoter within “the image . Histogram can be earily conm+tnyeted ab & initializes all histog nam QNTAYy enh fon each pixel 1Ci5) With the jmog Mstoanmm cers tas Scanned with CamScanner82 © Whok jy cumulative hislogram 7 Crive the mecthomale al expre anion » How dooy ety. exprph ghoul Hy qualty ob gn imoge ° > Cumulative Hislogicam - A. cumulative hinkgnam is ee ee (ahi Q mapying thot counts Vek en is the cumvlotive numben oh obser Ql th He bins “fo He ppeeibied bin. .That He emuledive hivgnam Mi of ¢ histegnan My 'S dking d ah — . { Rees Methematica| expression d& whulative histogram: NS s m where, n=-tota) number of ey preay “ ks total numbers ofs hn Name ok +e imoge hiptogream gree uniborcn néder equalization cumulative hishgram Ib @ twaul} nature of tmoge . Cumvlehive density fn have a function cf vaiborom h 06 Useree wil) ahways ishgnem_. \s +e prucerr of how we ean <*T25) OF Cumulative hishartem Scanned with CamScanner finally , UE Ran AQY thet iy3 (0) What do you mean 'o image Mhencement> Mention Some approar her to enhanze the queuiy of Image. Angwer ;_ Pimage enhanzement :_ The taste goal of image enhanzerm i isto process the image sothed wecan view and O85 the vidual tmiorynation 7 (t tonteins with here Get clarity Page enhancement is rather subjective becuse 4 deperds strongly on specific Imtoremation the. use, h Irermage, ne OS Meat era from mage enhancement ig technique to process an image SO that the result is rnore suctable than the orgined image, fore Specitic application, Htc suitablness is upto each app’ cation Image enhancemen} widerg we din Nefe, H# Ttis the subareas of ie Precessiny Enhantement approaches. b spatial domain 2: Bre query domain Scanned with CamScannerSpatial dowvein Tecmiques are based on dirtect memvpoletion of pixel 1 an rege Firequeney domein:. Terhnquerane Reed on modifying the fourverr dreenstorem of an meg e These aarvceé Some en han evs tom bination ©+ Method coment ferhnique bene d ON 'varu Prom these categ order. } b) What do yov mean by contrast strep behing? with appropriate mathematiea! Cpuations explain how image enhantement is accom pushed Yy 1) contrast sinete ning mary level mapp Answer: - conimtast stretchine + igi con itas t Sinetehing isa \ ' © mole (mage enhante ment bechugee the p aliem Ps to 1™ prove the. conmast in Qn {ma 7, Gs by smetening, dhe rape of intensity. valve, if teu’ ; contwuns Mm an "mage to SPM a desired =a Scanned with CamScannerVitow bit plane Sieing 1s aecomplished4, Answer — Ing 1s wrnethod of Represent Bit plane suiein an Image with one orywore wits of the by ter use forr eath pirel--one can use only msB to re priex fhe pixel, which rveduches dhe oy nol gf tevel to loinare ime] €- fore era ple Ish se A22KZ,Rbit~ fYnage ond we Know fhat t can fake valves fre pixel valve fore 9 bi —— Ov? |Do| | 610 ol] jloo | lef vj rT Lm petween O toy - 5 | Binary “abe oO % E fs orgine! image [ofo} T t os 1 D ynsé Fig: bit plane suretneg. Scanned with CamScannerof values s His toqnaws sthatenieg is also Sctid ay, contrast stretehuy Methemeticel enGhiesin fore ewninas t Shtete? x - Tun 8= (@- ming (greene, Pin) +Imnj By defaul} ver fore stondaged Agi seee Drage of Pht. TIminzo . Tmax = 25% “Se 255 4 Ao Amin Fmax —Amin where A> cunrem pixel intensity value . Hmin= Mnimym antensily yalve prresentiy whole imag e : mag = maximum tro tensity valve presend while heres: Twin Tmax Scanned with CamScanner© What ane edger inven brag € 2 Shows that the ? 7) forcy Pegnvorthoods of a lapleden Spercdort tan be represented Kannel! d [8 4 * © ~10 Solu tion ;- Edgerin animaget ny, edge ean be defined aso se} of coniig ous pixel position where an abrupt change of intensity valves occur: Edge re present boundaries between objects and baclgnound : Laplacian operratores 1s also & destrative operators whieh 15 used te Sind edger in an (ie The second ortderc dervvatve mask . laplacian is lso for. we & pees rr 4- neighborhood Japladen _ standard Mask .in wth eentert element tne element oF the op and nest ofall Elemens hare a shouvid be positive - All wormmere should be ze should be-4. NE ative 1a yse todake out in ward edge {nan image O =f OF tq A 0, -ly O pladan operator is Scanned with CamScannerwhen we “pply negative laplatian opencctors onan image then’we hare toadd the nesuitanp ; do orginal Image +ogel the Sharepened image on a Od J Pp Fags Scanned with CamScannera What arte neighbarchesd operations 2 faes ~ @ Neighberchord opereations sreply_ om larger — neighbore heed of pixels than point apertatins. @ jt Is moda a rectangle are und A cen yr] pyre ; @ Pr. SIZED reckngle and ane shape gi Here ane possle. . ®@ Fore each pixe| in the Ortigin mage he owteeme is umittere on dhe same loca ton at the tanger yn ‘ WNDL Ossian Scanned with CamScannerBlind deeomvelatim algerddim ca be used? © Meeltvel when yo tn bercmation about J the dis bron » Peemblead Gen hon resbe MES the Ange amd dhe ese stinlfa " cea by ust, om tartare, preeess a gamle bh the ceemlerctad Mee an" wetial guess at the sige of te ese 7 OPS spetby an anes of 1S as ho midi psp $0 blind deempeladen i mn 2 : perconits reReoveryy of the. ie Gel het keer smale ore Kucere Vane Sin the ‘Scanned with CamScannerpresente. of 4 porch deloreninad Or Qalene un pont spread genedion, s Wiener Fitker 2 Th is the, MSE Stadio fa Wweere Gdere fer mage. degraded by additive. nese, amd Llarereing . Caleslatin Of the woimerc —filtere requires the. ast mien that bie Sapa! nd ne Ie Ma eessesy ane ° seemd — oredere. Stati mare ; W Gefa) 8 Gty S00 Gady \ EG py San Cif) +Sna (e495 where , Sun Gur hy = Pees spesorn f ertvginal Sun (be $9 3 addite nese, # Gods) 2 planting f5 Here. tere, Cilfere nek oly perferems the deen Vola Gan by “yn verts-e_ Bictercina, ba alse Heymores dhe nase with Qamprte ssa opastadian - . Scanned with CamScannerS & : The Oey (evel s Sore bold mereed O% the mage 15 sheum “in Fig $65) 23 ablere conwlving. th by the kerenel -a5 given m $2 26 ig gs a0 1o AZ i 6% 6 ai Lo 460 4 6S ay en 35 (Se May GE 62 an ae bE US -L oO Lb We kmew 3 “ fy 2 Zz, Wk De Gi) t=> P= GOVE > Conas) Abo Cx) CCORICE DE Xe 85 2 2x52) A (oxsa) (2x13) & Cray) sorry Coot Cay (ot by OF = (A) + (ory (249) + Crxeg seat pas (72x19) + Covet) + ey ~ 94 a = (xa + QnA (49 + (vx64) = loo. Scanned with CamScanner6 Cs): Exedare rein, length encoding as em eompicessi om — fee, que’ Bngvercy Run ‘tengdh emending | RLE toa fe cani que used ty creduee the ovo, of a. reepeation =F ehanacters, This rrepeatin strong * ealled Arte Ayererls ALE ancedgy a reer of symbols ‘wb wo byley a count ant 4% camel . Ve PUN LENUTHE ENCaDIME, FOR. Qsmprtesst an sdivy fh 00 © 6 oTllllo2[ 000212} ©0000 Il wel p0e0 jo} Se TST eT O UtT 2&5) 9) Co OD ©) ty @I) Coy a a S Hy FS £L F > \og 58 et fos VPesleg sd a AL = (0 1) CL In) (6,012) Woy) (0-18) (t. 1) (7, or) (10°) Ena codes 061112198100) 0191 02/900 Loo] Scanned with CamScanner$) ! Expliim in details hows — Hasfiera rey encoding, » established t What wll be. the effe ctf tmige compnestm tsrn 4 aga an ordig, ig dh-e prcobabl Tely & Srerey wel wg qual 2 Priswe ree Ta establish the Human ede, We preteed as follow , A. Deleomine the prebel'th Ses gd. form bina tree by adding probabilities ture ok dime jalangs fakrn the tu lowest warilable vale) 3. assign 9 amd 4A ercbitriact ly to each brane ow the dryer Grea ‘ubs Apex, 4. Red dhe dawn . Here, symb] probalidty 6° 293 ood Oo ° Od oF Alo 0: 9 o7 ak 0108 vio) ens er-o 000 Scanned with CamScannert assigning. OS ond 4-5 to dhe breanches 0-5 SS 6 *esh Ord eo o'0b\ — TE oO tess gourd owl \ ( OND 8g gtd — YF aN \ 6 t o1oVaR eos 8 \ g00§ orey (1008 | fiad dha ede Gran the fp dun » TL Vo Spy bo) Westman onde nua of tidy 6 0 L | lo uo L Wes “ 5 yo “ g wie a b WAto Ss G Scanned with CamScannerflrerage n rm berc of bls one ppel as an expe ted value. abo 6) 9 [071d 4 (OHH) a oens)4 eo YY + easx 4) (0 ogtye) t{ oss) = avt TS dhe erebabi bra of every pre wm egal wlth while, MMC Comprenjer dherte He beng Gad of users verte able langth ede 7 -Tm Seth 4 astnprgo- sign writ be ital, an TE dhe preety fs equel ae. L We & : a “mn walwe_ tg Mahe than beferee_. ZAR 42K) FAK) A OXIA oy ROK) Koy SVS stit hy + orth - 92 trek t> overdo tam 1. S17 Scanned with CamScanner©) why oct is used rather than €FT as empge Ton NPEG 2220 . A . Nswer:. he key to the TPECy baseline Sompscession Process is amathe mehteal tramsforedien lrowa og the Dis , ‘ Hane Cosine Reanslonm (Der) “The Det Is ja acess alle met} cal openalbns thet ineluder she well kenawry mt Fourien “Thansdonm (FFT) as well a3 tram XS + The baste Purpose o}hese openations 18 to talee 2 stanal and -nanslonm + from one Type of nepresentedion to anethen . (2) What is monpho log lead image processing 2 How mo Faleal techniques ane established {9 irmge Pro cess Answens- Moxpholgg lea tmage Processi9g Is. a collection! non-l'nean openaHons elated +o the shape on tecduses Palin of in an brag, . monghokgice) ie mS can aleo be abelied to mer scale MSs such tha} +heln lighd sreairs fen dunchiony ARE Unknown are their absolute pixel values ae ot no on mimon interest. - onnsorvernag mare pve! ix, cleo. im Scanned with CamScannerMonphs lagica| techniques probe an ivy “uh Asmall shape called cx shruclun ing ele men Monpholegtea| openatto e1 O-= how rtalso off = background Pixel I= hgh we ron <3 Foreground pixe (, enphobaice| “echolques ane elabl ished mage Processing 1 Tn image Processing mathemehica\ rrp ho logy RSC HaSIOUmtmerrnt els iden ify and ertsac meantighd age descniptons basel on Properties en within image | i" 9 Distinguish belween monpholegical Willing end ing When ase te placel Ina bin I Leach of the Pixels tn the SE 4s ossociale with 4 i Cowessonling Plrel of neighbour head urdeg struchuning element, Fring ‘ 1 ' uy, 8S, SE Hs seid to de} cin innge th, bnendy G Ws Peds tat 'sset 404, the connespondag mage Pike] is also 4. Hitting 5 said The SE 13 said to ff 07 intersech an Tmage i, tan any 04 ifs pdels thal js Sth to 4, the corresponding imare wivol te le. a Scanned with CamScanner© a 9 ~ \o — 43 “Nolo © 1 Of|- P a a oO © |= 7——| e[_ - = ars 7 tye [oJo|o[o Jo in oO O/9 | Cc Scanned with CamScannere 2) Foun ten Transhorm oN ) The Fourcien Treensfonm is 2 1m porchent tynegpe Processing too] which fs used . to is wompore an Ymagre into Hig sing andessine Components , the -ormuk omd mathemeltical subjectsot- Ttounten tetanstonen can get que cone Hated . 150, FT can be ae Vievally uich simple - G1 can be Visualteed such that: 9 blue brown Ine bpeale down +h ET inh a reality MoH ye Vi'suol system {s eesy to Follow wayelek ts anothen. exer fe by which FT can be visually analysed Scanned with CamScannerDY) Answer me Fon a 2p comlinuous tunebion Seo) vs a ot onelen ( (P49 isdetined oe nef J nity * (my) chy dy Poe Mae Dil 2... Sonvergely, "™pa) } uniquely dedemt nes Fey) “ae eects ane delined as. Loa)" heey Indy w Se M19 R= ry =e et th Stn yi eT ™, er mm) '* rgital | Image , on Mog 2 zs Ge) Gy)? Fes) > 2; 7 ton) oa oe Ao = oe u-%) QA) ks) = m,~ Ke Re Cr) me) Mis = RF one 9) Fo) Proved Bul os Scanned with CamScannerCc) soledions Moment Invantent, * DMoj= 2 : wo Wy! (deny) ae a Fe Y dey) = IXO+ 256 4 3BxS4U ke LIKL ECKL FI KEL HY thoxry loxg = 2uo Again, mM oe iM Zz, 3. ; ny 4 (ny) = ID) Simanky, in Mo,7> hs 6 20 %, Se Vv Fey) - s 10 2 net Rd * Hau) = (lke) Uk e) + Cabte) +x) (sx 4) L tExs) aay s\ 4 (6x4) +6.) Mux o) = 14bS Scanned with CamScanner