\t-4 SINTRODUCTION +o THAGE PROCESSING tom. i: =D function Flo4y) 1a) (plane) Coordinates | Pigital Sage contains diffecent types of efements e fietuore _elesments ,pefs 4 __ OF ; 8 E [2 Pixel ie the teram used ornost widely 4+ dernote Ha dlesene nts a} gical feaninge. oi z = o my ee . Thew€ Onre “no clears cit bousn dosries ian =the coro nus Processing at one _emd -to_cv of Other end —_ — pS —5 types of Computerized porocesces in 4 #. Peerigeeet apemcecce, [ hl 1)BA lous Level plead Bicone Eoceece ay tne fou | thot both its ‘nnputs and _oukputs._asce ot Mid _level__porocess. ea _- | 2 94 Smnvolves task. ‘ 5 Syusenie seconenastiene Sem ony de sewiption of these ix jecks tp are, , thers +0 foren suitable fox Corn puter frre cess 2en al otassifieakion — ofp individual objects. — % 9+ is chew askeorized by —that_-fack thal its | fan pants —_genneacatlay — Aare ath bubes extras bub ide dubput: _foroen tose lyn Hig hk ley £ rrocessey — =—— ° St involves nf ae ore. ized — objects, as_asn_ tomege _ornalysi sis Bob the for of He Coroners pex-forcmn ian iV j—saexsmatty — associated — cae, INPUT NOVTPUT _ o A ano IMAGE ESCRIPTIDN OVERL BERING e | Diqerat a EM AGE! ace SOpBEUTER | cay __| PRocessSiNG VESION MAGE __s PROCESSING J -eom-Pu Tere | DESCRIPTION Gpppric: ’ Ac a ; a Tenvolves —bapheriang arn imnage 7 © Cosme ara oor _Scano meas ore ‘19 cxletiong fnage trot. Cor putersNERALLY ARE ‘rmAge! ous Mowpholoaical Processing Seqementatin, BASE Represertahon osad pesc-ri ston, Object Recognition oe a) ah 4 = Ser. Bas Hasse — x oes Pdi ss ATTRIBUTES” Ladera ee 4 Hae original Out the hidden & Wa Wlight: certain mney be Pann pox-terrrt comtrwast adjustment . «brightness & 6 Trenage that haa of: escinn dae de gracedDate ) ane i ‘. 7 cence Sep chucles Of ont. TLLa_smpolenly Pe rctpenenaliat sn nine iene tS of Colonr aeys Pero cess =e |S this 4 of. colours eno & dome | 9+ is eo _prrecess of prroces Img _coloucred vineaea | eituey as RGB or BGR. Wavelets amd NutiiResolution Porosessiing_+= Wavelets are snail) waves of United duratio: | whith axe used +e calculate waveleb srromsfor am which provide time frequen leo formamn ation of oon caaeaee wavelels lead to Nutt resolution poroce. Bsn in which an im is wepresented tn of Fe ie Pecfoem aiff” i. q operations) Com poression {= ay 94 deals itis techosiguiee fer savinng stor: space mwequired +to_save ee ae o || $+ heals 4 antsSegmentation Sie sof, “pacesti Homing am inne ge : xemeaeeete tet is enerre tly used +o lecate an object & _ boundaries of oon Object: esen-t+aheon and Desc~wiption ci - — —€—-__ |... 4. pean Imie a Sut-table fre corm puter St bas sacri oo eS —— Types ee... exrtchdn Regional Represt® His used wien fous fs om Ite +-nal Pee ica URS aextuse inne Desewiphion deals with oo an attri bute these Sthibutes wesutts Im sdhne quamdtter i subi ore of imdacest + —: Sane SS Thuse attributes ave used +o Oi ffecentiake | between ne sek~ - ject Recognition +- A OSS prrecess thet assigns a lable ty am object am 3 es, : {+ asstons [+ car. - pa mewitdge abouk a blew S is ¢emonowledae base: Sela info” ccelated to tenage press stored iy nontows lect ac bose. ——s Temage a ° x © ae ) Mostly the ; fosrom of ay el es Blokes ge Li we cam ot app digi a: dso « apie. Sienna din d A } : a So__we meed +0 converct tu comtmous | sensed oclafa (oa st eae Mette por ~n fore ns 4 i ~ a eH Amd +nis. ¢ meee ° ep Ser pling -& quantization rcV ane Comino, ——> % Ky jane Los ° digitiaing co-ordinates values ee pling ae ee Wr Onis Ha we dtaitize & 19 is dene om Indep! sadewnt vorlable . Be owe BSume vowiations tm ne sample si. gr which axe vandom due +e moise - WISE OF ia # WHITE PART Aico So Hea Casve wrepexaaent’ VarioFes DARK PART SAMPLING ees -eomptitude Value. Om plitude tats ee ~ | is based om & poor-bitions cl i eat 4 - NORMAL impor DIGITAL Lt AGE TRE Be PRTER SAMPLING be QUANTIZATION SPM PLING Qun rnp zation. a. ae 4 \Comvession of continous Trmage Into a ola - “seFof points by selechy Cermvest Mom of. Cortana Pinel tmtensity values fri: eer duscwrete Levels, postr. : Pe oA sk 9 , [eSt forrses on Intenshty, ya! spaHad imfo” fie, he Pasig hires Oe bolo ee location of pixels, | levete He pas. = . 1, fs dete+rnined ly v73+_ fs deterrmimed by 0 of quantization levels i.e. 0. of intensity Veckeatd Levels. pling oats pinta pyre wok te fovege abit “color depth for 5 mere Color varsaHions ble it is eee Of ter. ad tem is a a IMA ee HARDWARE SOFTWARE i= E PROCESS IMAGE PROCESSING eR Bei ; “s SENSORS FR RBLEM pomarnnyCommpusbesr 2 est Is O_qemnecral puis pote —carcnpubere tobeicle ie tsed in, mo fe. ay Tomage age__processing se Seftuonee + Pic. khearerermmy £ gore sed fan femnage —poca.cessiong a oa oh Mass Storage + Se oe eee 94 Stonre s seers toepinete. of Tengen dating ta ae ee pen olive, exteynap ROM device Eomengs Diss Display + v é a 2/94 imctudes _-Hae “mem tmer ovr Display Serecs , ae | ; “3is +ne commection of oll the above —— ments of the frmnage —pocacessizng- sy stern. TMA GE SENSING AND (Yor WIE ee a 4 —— egitim is the 151 Step in Die. elements _Orre ee = ocndl acquisich’en oe. nn. ee rt Jomage _sennserrat a 3y Lime Sem Ser. _ 2) Ansa Semsore Sols pba oe ye ° e ge Acquisitown using ing Sea J ENERGY peer Well» ZZ es KEREEA |ut Semage Sensor used im Aaitel carmerras is devic, shah converts om optical ines +o __am electric — - Inmagqe . Ee Tusse Sensors Sense ene4y = And filter ftlterds all the unwanted Moise Lond csadi atoms sensing cmmoteria) Senses the ener amd Output Hts energy im fooemop voltage. : OR. | COMPLENEN TAR, METAL OXIDE SEMICONDUCTOR - Vigne Sitters that favou Pes siremgere of pata r —> UNEAR Hotion Te gene wakes 2D Frage using sing bs steve Should be_crelative “displace ents ix bet wy oocis. blo te censor B anca to Lb. Simagedl . Single sensor —* Mounted om a tead Screws of Sty Se Vv — providus motion in pex x! ; dike Dion Pespendhe “Tala sce) ee ae +901 ( ! Vieneos.—olanecbions , _ Tusae Sub itizex ced Miowze enae E ——— nets \ aLime sensor Cg oR 5 LINEAR SENSOR enged fra forrsm of Sensor strip provides tt ging _elermecnts ing 1. hinse chow. otiown 1 to strip “provides —femaagiong ion othee Fi ——————_— are HNE Horio | SENSOR STRIP JE SENSOR “CRCULAR SENSOR org D Sinn g ore aresr ays 5 — LAVAL£ te, ’ Bon tats a of nla Gr seg sone ae 2 4 its “adduacrdage is 4, | complete imnage cam be —nbinined by fe. HH enegy pater ones the sucnface Bec i Moto is not mnecessaxy ton Hts Response af each sensor Ig % oes eo t ene i He Sw Semso--. a $4 is cormmely used tn digital Comers - bof ced oswary 9ensor is most Corenren bony Bled Se-, ———— eee. 1m tHheere Com eras Digi . es ht tect oe : a bel Lek im ae ee at cerntisanup.—vartia bles Rb. ofS Z. it Suppose that we Sanmople the_coetinous irbasic. Woys ty et present f(x,y) urn Chom covrtainn 2 axis +nak = Tod Location & 374 onis — - Values of f Cimtensity) al me sels Bas) eA Eee Twa is more Commonly, used le: 2D wary. ~*9t Shows ftx,y) as 1H would appeax on a monte, 44h tne inntensi is mnoxv-mal'zed to ‘hntexval sa ©,1] then each point fxn txnage haa Value ©,0:54]1: iis. “d je Tris wep-resen+talion is simple +o display tne ~numerica volu the mu-merical values oA ono ornate x).oC Ts wreprres exriation | fs UsePrl whi, develops algonictwrers ai =Fte, o> — Flo.) -- Flo, N=!) ~ gle) Fe PGND + tmay) = fa \ ‘ 1 L He to) fleet) ELEN d) oe . Cok. Y. bean etement 2 omabaix is called pixe), = ieee demene Oe pichire elemes} = 4-+- oy Ovi fin of a digital ¢ rari ‘is ak top ep Tee. fe,0)+ 5 Pl There ane mo crestetobions om M_& N 4 othe~ -tha-n__that they Should be +Vve: a i de z . a oT due + stoma & optinsiagg haawdware Comsidexations -the sno. of Utensity Levels a typically fs _an “thitegete power of 2) | =] by|| Sabuwation fs tne ighut vatue bee pobi: vg aU imtemsity levels cue clipped (void) J 4) relge appro Sk geraicny tenure potier = _—z? | corog Somoge 4 2 : | See — ——‘y/Colline: ‘is. at pniiteeaeat de cepinaa Ladle da Sixenpl'f'« ‘i sbjeck Ides ficahion 2 ectachon frm o. scene. _ da gr ts divided ‘tote 2 -mays, =a Gt Se Scot /by) Chava achewisHic s _<_—<—_| => : ae eee say Radtawnce Tock arn ount_o. ener Reece || —Twatts —w). + . Spee of ge {= b)- Luominancce : Measucre of ocmnoumnt ef ene Lumens = "Io ) On oloSe-svex pewcetyes oven — Sousce ._ | Ahahtnes suent ~J-1_ fs subjedive mmeasusre | (Ho unit) that 4 bs t \s_prrack’cably thon possi ble + measure eae Red L _(435 (8 bm) (SY¥6. im) “Geonrn) Secondany "Solow = Magneta Cyan Yelouo en (+8) (R+4) i a oo seepetaertatice ead class Oy Th-res f RGB CMY, x¥2. ] — - 1 8qR. e by Luminance be ee HSI, YOu cr mn Hue—e- value of pixel befix = a weet Rocersiy 7 s>-vaiar prxet—ocftere TS Tem sformahon = “a pus clon Processing = = t 2 aot opoxabes om single pixels of Ss amneniputation & rmage thwrese 7 Tt Vp intensity Leveles? “¢ , By recon [Negative RE Tdentity] Bea 2 aos acre Lleg & amverse looy. Negetive Toramsform ahon * Negative of can *Y na, ae eq ahve eo - = ; S= b-i-+ | Level ais ES wt ts suitable ae emhom cing UWWhihe oy —— Se cited Be dino ps | = ae si = toler ox fntensity A i Ot 4 axndidro —feenage— ap fan Ion p. a Levels —{an_sramnge_[° 7 tat] is obtat: —_ bY Tdentity tromsfoemabion 2 Sota ore 2 soteiity Eser ep “the Tomage_af Leamnsenission fo tdemtical - toto} ___ ——of—_the =o renege _ a ep vw Each value of fon puk ¢ “mapped tp each othe: nc Pte np Y= im Serie 4 | ee Terems-for-mna toms =F SS oY eg_crrematerm ation: tt bo ae in sghee Soe ee eT Wi ghee intensity valises odeswrUsing this expression wwe Carq corm press ox aes leved . Wii Power Law Tromsfor-m ation ———— ere we USE gamma _—tbromStormaton where we need y | +0 expand ov Coren preess derek er ceegion el isi = Ga ei where ck ¢ orca constants . ep 7i 2 x power “ys 07h cot - - Tata ee ae ‘Ip are levels. ok° . 5 ie et mohappim tg limeose (ole faut wt _w Shey yt capping st uasiyy a dake pulpit “aber ————— Fo ris = ee rr s 1 ee a —— |_ Hi stegeraxm = - ZN " - =e Oo] SE FS preecese fn _toblich w0e_prea ets m= : = 7 ol image tp beast ft cre a 2p SO Ts tne tH gray level Be 7,7 Te oe of pixels with g-ray Level te then ite ee is we Tree.) = i Ee es i HI Sto. ae ie alten a Sis a _secnnfque of amevainl a tdsingoen eae Cet ain pottesms :=. E . Pee | FoR PARK IMAGE? a ag AU the valitet owe om Lapp To _vhhamd. side f ra ae | = eae al Ee —— 1RiGHt HAGE? FoR. BRIGt q es Ail volucs arse seseryt on Right ibud Side. (6) FoR Low INTENSITY IMAGES + + a . 25s I Al Velues are present oF : j sapaate = 3! (A) FoR HIGH INTENSITY IMAGES:a jes epeoneen F Coy Fs - Gy = Se . aes 55 mits mn eth ee ale rea fooilipdin fe fe S=1 {Make a frequenty able & coum fre — of each Ble ne re} Im above -euisians iter.) oe () sues ie ee a (sobe) > ee a4 ane 2 ae by A filter fs enUEA mem = Unto ite Outputs is mot ao VUieneose Cor Yi Of He foo pucksee eq e's = zoo 2ip222252 249 2s 126424 eas) | (ey wens Filtess 2 ee persis __each pixel “vote —wi-Ha the medion i} — _ pixel values ion ad = Meshomics. of spatial filtering. Ico, 1] C0 9) CO; we] Ww. es eC Elx-b (A=1 eee (2 . = [utls Ye H _ten_tnaage ae BN EN e mask oF | size Mx»_ 15 || Te es a grey level fox ony | Coordinates gimp = ene W(S,4) f(t, yr t S=-0 t=-b ntl ererleteme la) ala b =ln—)/2 Steers... Mi) 9 eens = eae 1Ye The Wolves (one ficfent) of ste —sonosk. dake, == sHae motire oma! — Reeere EE Snocntiing a Spotia) Correll a Most 2 Corre lation 19 commection blur 200° mop things measure 94, com pute: creer of_stvnt lari Hee of tuo9 F a =e fp si Saal as “Hiey oe shipted wita one pant = |. commen BE of APE tes 9 ) of size —__ || wots ee ari —_ftoey) oleate dO vo (xy) pee : _ a WS (9074) FC aPz = = ; ws, t)_ Fest Ss-a ml ES pata comers Se oh = a. oF *|| the pinay operation isn spatiot Hi ltesiny — whese Samant fa__applied +o an i + | Comvyolutiown op 9 filter vl 5y Tze cone With om a Flay) denoted as wl) SealeFousiew TsromSfow-m, amd its Porope-s-ties = 0 Re ce ro ° sTextince — Asnalysis.. ———— Eno Haat ohtouos Ua_ctp are pore Seat a ie Cea. = Sin toe feu ew doxnain instead of Spott Slo ois. $m +nis we see an nares Os Colle cio of frre quentic | imstead of pixels . | S4+ is _used +o decompose con —famage inde ‘te sine & cosine Compomnersts..S| ea =cnalyst's SS ae a Ss is breakdowns fete its _¢ = So “Mat we Can_ amalyze cen dl Unde tang — EE - ne ro oH} RECOM IHVES S50 — ar JL ae aomal = : sae sere storm _clomnalins See ‘ona ta elie | —— Ee et Ui) Cosy re Task. -{ ee tara fore den ain + —=Ss a Ui) Apeuy t inverse -bramsfeam to oe || eeteatitel gin ian, —— 1} Poco pers tie a = a e —_}l 11) Lisa@ecr-by = PT ts a Umea operation —_ as Ui) sip} (Troms tation) —— shiftin results ‘noe shi in —Pprequency —___+— (ih) Scaiiong — 94 ofhects =e gh sera fay — ‘aortic LUN) Syren ety = Fore =real voled ~? ere ee Cece (v)_Poreseval’s Toren the energy of Loe | is ae — Wi) tornvelution aThOWe ey = Convelution fap ti! — lorena CESS Spt tn Sy Poe Me So _pareg,, Lomaln . pa Wii) Duodity flM— Flu) _ F(a) — $(-a) ce os fet: 4 Applications + Somage filtering.» sconage Comm prressi : SO eaelloe mage Prnohysis, Weise Reduchem ce - “ - Sa uency hormone tees 6 ti These ff |+enrs opewote om the Fourie -bremnsfe Of Sg =be. codify its forequenty ° e ° > sail pectin effectively atten eA fgg Spats _Caewa chert stice — fan =the Hime dowmnousa..—___ Gi) leur Poss filters (LPF): To~remove high fre. eT, te) ° ° __temponents Cmotse, time detaits) & retatn” Low c4y__Co-m pomencrts Gi) Bomnd Pass _ 5 fi raemacs = ere as eS Steps® = =—= ~ a © Taran Sfoscen Joon ————— SS ee Ree oGHRYORE Se fetes teers a _> RES | Noise Recaichon, &: Applicahons 3 — ___| Aetection 4 Reahiree | extrackow. bl 9+ is a Peatirse extira im 4 ecnoly sis 4 Cv omnd 1 ub S+& primacy — puscpese ts 4p olebeck Simple , a Shapes Such as times, cineclea _— ——— t me ts el — Paasemety!= facrarmetenr Space > ~ ee ee eet end. —acgein slew ae ae fe Space th —cureves or _susrfeades fn a —— co Biren bei = SSA SIE is lef aeused +p. commneck — Se ee uation of 8 cae a teen re — ei =e mmMnat+C om = slope C— intercept of the Vine. 2 - ’ — ‘Space. z x = Se een ae nae, plocne—_ Jooto O. imne imma tne Cm plome ed = > | ‘ ES Cs) = Curry) 1 Seon i — ee eee | ere NN a il Pecrarmeter ¢ ow Hough Space [sp Ake Barre +00 poiarts commected by a diene no wtag SpoHel —do-meinnq—ctety— L3THL_oe fn-temse chi Uimes inp =the. Hossain Space. ae ~_ een ° Edge dehectiow ° Ee detection . Mapping eclge potets e emt fy local amaxi-ma in : Space aS s Mapping Huse smaxima Detecting peaks fn the | back ImageOe 2 eee ° = eee compufeatomal ik |* Shape alebecHion — Pe & age sereaen i, ee * Shope re cogent * Medical Janagi ~ — + Dbjeck es _ _—— | age__Nofse avid _RestrraHerns — ae (tt tomage noise —is_omoy usnwanted sromnden | Vasa Hern faq asa Hebei tae ss ae.8 ee “rena ge acquisition — Process. Noise occusss olue to — © Se-n Sone : Dota _leenipeession Pi 5 ae= aes HH stimulates” ES eicteiteisles Siena th _| | 4) i, ee ° ee Pope - P(z) gee” 2 Pt) —_ =| 2) Jes a OSES SS gray level SS mean of aug of 2 a > Stondard deviaHen = > oo" — varriamce . < a I 2 ae A $lece Shand Pla = 7waitin A. =~ Neem 0 S6 Se Eee SW ees aa i} pPi~)SST oa aaa aa ea eT ——" L Median Fi ters asnilithalnetiba Replaces each _filtes / pied Uta _-the -median ee a t-ts oni | oe a > Band Filter 14 Remove pend cite ‘tampa b Acla pHive Filter! sse~move f i. Recluce tatpnsction 4 preovide S$ a =