Download Module 1 - Lecture Notes and more Study notes Digital Image Processing in PDF only on Docsity!
MODULE 2 UNIT-1! Digital image fundmentals ¥ Digitat Image Pret deacke be ii ‘to pmcning Qo two olistontopat cture by a cligital Cornputer, * A monousrome 1% € 8&8 an image bh a two dimensional intensity function 40uy), hoheve % ancy denote Spatial co-ordinatin and tke Value 3 t0Ly) & Psopertional to tha brightnians (or grey level) i] an Image ak (x,y) Point. ortg ing y 44) a * A digital image wan image 410y) thot has been discretized beth [n Spetiok Co-ordinates anck beightness, : * A Hagital Image Can be Considlered aL a motsic Whose sO, and Columns Inclice, yelex to a point in che image and Corres pondin a matrix, element value rent identifies the Grsouy level at that Point, * Metrix Clements digital image matsin ave "| Colled as image elements, picture elements, plats os pels, Types 4 Emaae, browsing * Low devel Prouxing Include, notse veduction, Contrast enhancement and fmage Shaspening. Input and output 3 tow level Pamaming ase nae + Mid Level processing Incucles segmentation Prous, bshich os | Prroters, posstitioning an image Into regions. os subdiVvittons. Here inpubi axe mage but outpub ave attributis Extratiid from image like edaes, Contours cuncl identijying 4 individual Object, | oe High level Proussing Involves fmaqe an th, ( uNnderktatin 9 4% Collection 4 ecogniaed objeck) ea '- In atttomoted Uassitication feuik and vegetables Images gq ou Kind g fruit and vege- -tobles owe abyutred, Preprovened “(tow ever), Seam, - ented and described (mid Level) in a form Suitable fox Cornputers Paeeiieas and finally frutt and Vegetoblers are x nized Che veq *cog ! C ‘gh level), > a ° *~ Fundamental Ateps in Rigitas Lmage Paptessing Problem Domain: Image or data should be taken etsoretiy to the weyutserment, ox applications, applicable to other type 4 beige bai biomedical. images * Inge sertoration:- Thin 4b oa prows reconstsuctior] ot weenie an image tot hod been rade by using Some prior knowledge y De degradation Phenomenon, €4!- Removal f Tr alla * color image Protessing:- This oma impostanie due to use canin Stantticant Increase fn" te digital images over tk, Intesnet, Jhis can be divided +o two major areag —> pull Color proussing }- proursing Ha Image e id Coles acnyired with a fut-Colov senor like*TV Comer, <> Pseudo - Color prousing!- Assignin porticulay monocrsome Intensity intensities, 9 a color toa mn Waage F * Wavelets and muttrrerolutfon Processing '- Waveli ase tte foundation tor Bepserenting mages tn Vortows degrees 3 Beolution . They ave also wii used for image Compression and nolre elimination, * Corapre sion t- This Atep Involves vecluthn the Htowage Spo veoyuired and abo tke bond woiolth peoyuired tov teansmitting, He fraoae stowage technology how [mproved Aignifican and BW Copobi ity has abo improved a tot due to te use optical fllowe but Si image Compresion is an important 3tep and dhe Atondawd Ctik- -narne..3pq) widely use. (aPeq + Jofnt Photographic Expest Group] y> | bs 4 Morphologicat prowsing.- This Step involve extra- ee Oe -cting [mage Components uweful in the veprerentation and description g tke Shape, x Seqmentation ' partitioning or subdividing tia, Image Into {ts Constituent powts os objects bs Calli as, Segmentation. A good Reqmentation protecuve helps In Sucumful fdentification 9 objects in an image and a weak Seqrnentation proteduse may lead to foiluse. Thevefove Segmentation pron should be very accurate . ~ Representation and chscsiption :- The output Segmentotion pxrows by wal data. Thit data howd be propesly wepresen tet In order to help further Prouusing. There ase two types 9 sepsesentation —> Boundasy representation :- Helps In studying external hope g on object, 1 => Regionot seporentation:- Helps In studying Intesnol pwoperties Uke texture, > In some applications’ both types 9 Cepsetentation B ued, Deseviption os feature Selection deals itt, extracting attsibulis foorn wepserentation bshich leads to getting Some dwuahttitotive information| intewat or helps in ea penti ating one chase 4 objeck from other. ‘Scanned with CamScanner i physical dovie that w& Senitive to a band g @ Letrornag neti C Visible, Infrawd etc, and poduud an electrical eng: y Spectrum, ike X Rays , Signal proportional to enusgy level Sersoat > A Roleiser tor ecinven ing tte device output to digi tal, form. Network e Compulir "plows ime i i Mass Stomae Hardcopy figuee.2 f image . Preedthode Troge Sensors, peoblun domain + Specialized Image Paowising This Consirts Performs some Operations - tic logic unit and logical Operations on , Auth Opesation Os trey ave, digitised nose, Thils haxdware i. Aome-tim v hawd wave *- Dawe Wave digitizes Gnd a hardware that (ALU) bshich peyjorns Lusithemotic Carte may be axsithina- tke entive image, one is Averaging g whe image G2 Loon, end Subsystem . Important CharackeyiZtic. ters device ih ft, speed Q Opewotion. in ovdes to reduce tte 4 Catted ag front “% Computer :- Genevol Purpose Compubes is sed 2 image Prowsing System, It ton be Fangedk teom Pessonal Conmputes (pc) to Alpers Computes, Sometimes Specially dusignect Computess ave alo used [n oxen to gee seyulred level g pesformanc, + pelheont Trage Processing uses Speclalixed son trove foe Pexjprming Specific totks, Some wetl dufgned. Aoptwoawes cLLows the Uses to write Cocker, £q:-Matleb,c, Astra. Image » PC boved “i Image Prowsing CREAM— usec for biologicot image. GRASS > Gjeogsaphic Resource Anatipis Support Sytem, OCTAVE > Open Source —> unitx Plat {osm treo image > open Sourte for pwject cevelopess, * Mags starags Mass, storage, Capability wb te most lo portant Pors t g image Procenting a@pplins An image g Sie 1024x1084 pinals venubrea one megabyte g storage Space D} @& number | images ave used then elidel’ ie becomes A problem. These awe three types 3 sabionamge, usec fn digi tol Image. Prowssing applications, i) Short tesm patasttng - used dieting Processing ii) on- Une Ftoraqe, - fos fast (foeayisen tly Used) ti Axchtval 4 @- fos storing image tohich i> d ase not dredhanl Used , Tage formation In the Eye i- In an osdina photo aphic Camera , Ae nan has a fixed focal length, anct focusing ot Vartous distanns & achieved by veaying the Aitance between the Lenk and the imaging Plane, ishtite: is. Pilm & tocatec. In tee human eye, tke Converse ik tours te aistance between He Lee and use imaging tegion (she, retina) & fixed, ancl 6% focal Len reeked to achieve proper focus i obtained by Varying ta Ahape Q tte lens. The fibexr fn te Ciliary hooky accomplish this, tlattening ov itt tox. “ening tt lLnx for distant @ neay Objects, Berpectiv. ‘ The dittance between the Cones te Ln and ta setina along the visual aris th OpprOxi mately ltr, The Younge 4 tocol dangths is OPPO rrncely l4mm to limm, the tattey taking Place when txs } 1 #rom—> The above tiquee (Uustrates how to Obtatn Aimer’iong 4 an Image fovmed on £e ting i Loa h denot het nt 4g dat object 100 1% In Sh, Felling image, h= Q-Srorn “te. Brightness, Adaptation and Discatrnedét ination on | ¢ The Mange light Intensity level, i-eur important | to whith th human Visual tem can adapt ff Cnopmous - On the order 10°. #I0m the ACotopic treerhold to the glave Umit, Experimental evickenee indicals Mat subjective befg htrex Cfntenstty as Percbévedt by the human visual System) & a logan. - thmic function Ma tight Intensity Incident on the ae Fg ae, cu plot 5 Ught intensity Versus Subjec- “tive brightness , (Uustrates thts Chava cari Atics, Glare Umit - 3 RS (gle 32 dig 0 uy Suotopic {] arcaholt WY log 3 lrntansity The tong Jolid Cusve sepresents ta, van € QD intensitieg to Whith the visual Suaten Can Adapt. In Protops « IC Viston alone, the range ts About 10% The transition trom Phetopin to #EAESBIA 4cotopic to Photopic Vis~on W qroderad over Le approximate Fange feom 0-001 anc 041 millilambert (-3 to -ImL fn SMe tog scot), a tt, double Pranches 9 the adap. -tation Ctitrve in thts Fonge Shoo, Two phenomenon Uarsly demonttralis thet percatved befahtret, th not @ Simple function g intensity. The Jtxst 6 based on the fact that um Vigual Syste tends, to Underthoot @ overshoot asouncl the bounderry 3 regions 4 dijlerent intensities. Below Pique Ahows a example } thls Phenomenon, Actual, ? mtev ty we Math Rewctivet band eye intennity. The Colove thak tumant Percetve fa on objeck ae, dele vanined te neluve of the light aeflected From the object. A body Mak vef Luks Light elab ively. Lalanced én aN visible wosisalinats.s appeerh Utube ‘to the obseqver. ithe withouk Glor & Caled nono chromatic or achromatic Light. The wy feohure of menodrromalic Light &% intensity. She tntensity of monochvormatic ght perceived bo vary from black b fy ¢ fenally & white, She rm gray dh 5 ted easily Lo denote mene hronebve Enbeensity. Chromakie Ceolo1’) Aight Spans the @lecoo magnebic ntergy Spectrum from ap Proxim ably O43 £0 OF9uUm. Se three bane Quantities of Chromabie sig be one, a Hadiance : Tek th the btal amount of cnergy thak f lows from the light Source. Cwatts) by Auménance; GL Giv& a mehure of te amount of cakgy that (flew fer tc Liptet Semee) om observer Can perceive from a light Source. (unit - Lumens (Ln) e7 Brig Ener ¢ ft uh a Subjective desevip tor of light perception that Practically fon Possible bo i eabure. YALA AZAZAZ AIOAIACZ COGLARD TONOAG BAA Somage agqutsition iad a Single. Senkog- Example + Photo diodes ; Leg? cs A gaeen ED ofp will be Slronger por green light thaw a other Com ponents in the vizeple Specbrum. Anis th accomplished means of green Cpaxs) filter. in front of we Len Sento). ww) generate a 2D bmage uring a Single Sensor, there ob bo be relakive displacement: tm both hosiqautel GD ¢ Voubkical (y) dereckigus ‘beween Me Sonsor f Ma 472 bo be émaged. t felon (: Senior. G Ss Kothtion . son a K ———? ‘Ainear motion.» a One fmege lene ouk Per tnereoent of rotation and full Gnevemen® dinear displacement of Sasor from left & Tight. A fiben negakive th enounted onko « drum whose emechanteal Yotution Daovides displacement fm one dimenséon. A Single Seusor t% mounted om « load Screw the. Provides yuskion @n SE Perpendicular divectien. Advautang ds ° 17 leon expensive (fn Cripensise ) ay Derevi des ar i veSoluki¢n fenage. 4 Dibedweustage . Slow Precess, Other eoramples are , Miao demktlometers 1- A flak bed with the Aensor ewoving an boo lmear directions . Image Aegquésction wy Benhor Strips. In this arrangement the Bk ip Provides iain 2 te Clements im ome direction, Motion Derpendi cular Lo Abrip Parovicles sbenogiing in te Olter déiveckion. Nas type of Semhorh WL made up of Ho00 of more in ling hensorb. Other Applications ares ay Com prutentyed Oxtol Fomegraphy (cat] — Magnebie Resonance Emagéng Cmas] —? Positron Cm £S5 tor Tomgaap hy. Cer) Jrvage acyl action valng Senses arrays, Edementasy Senhorh ave A77%an ed in the form of a BD - array, leads to Senhoy arrays. Gg:4 Eleckoormaqnekfe g ultrasonic Sensin deuied. a Digital Cameras [ cep Aarau typically] Acres Semsorh Lypically heve 000% Ho00 Cbemrrts or more. Advantages 1 + Motion of Seton £3 mek mecersfany. # Nothe reduction £4 achieved fy Enkegaakiog 4p light: Segeral wath raguk Lo time. Digital triage cccguéattion proces Illumination (energy) FN _ EEEECECEEoe ee Output (digitized) image (Internal) image plane | Scene element * Tiumination Source Ppaopets + ed energy oma Boome erent ( obfeck & be rae Some Dart of enengy being reflected fom Scene omen. A Tenaga System Collects he veftected lt 4 from the object ¢ Joos tt onto an brage- plane. ustuith veh Colnetdent with the aA She Sensor avaay, the focal plume. of lenh , Dreduces op Proportional to dntegoal of the light zecetved at cath Sensor. # ohnadog 4 digital Clreuttry f Emaging System Produces as ¢ dig tal tage CSignals) vapectively. A dimple mage formation Model. Ay bemage &% denoted a @ dimensional. function of the for fy), The amplitude of f? ot ey th uM Positive Beadar qoukte, usbose value Lb | dekeymined ty the Source 1p fmage. (2 Devel lnktanity volue depends on Source of emage. in other words, intensity valu are Prrofortional to energy vedbluked oy the Source. fey) must %& non yoro ¢ finite. Le 0< flx,ye oo ([ftrel valuss ave mon ? $ finite) ae She function fos) th Characceriged ty 47 filuemination —» £ Ge, y) a oy Reflectance — ‘VCs, 4)
