Download Solutions Manual for Digital Image Processing 2 and Analysis Computer Vision and Image and more Study notes Biology in PDF only on Docsity!
Solutions Manual for Digital Image Processing
and Analysis Computer Vision and Image
Analysis, 4e by Scott Umbaugh (All Chapters)
Solutions for Chapter 1: Digital Image Processing and Analysis
- Digital image processing is also referred to as computer imaging and can be defined as the
acquisition and processing of visual information by computer. It can be divided into application
areas of computer vision and human vision; where in computer vision applications the end user
is a computer and in human vision applications the end user is a human. Image analysis ties these
two primary application areas together, and can be defined as the examination of image data to
solve a computer imaging problem. A computer vision system can be thought of as a deployed
image analysis system.
- In general, a computer vision system has an imaging device, such as a camera, and a computer
running analysis software to perform a desired task. Such as: A system to inspect parts on an
assembly line. A system to aid in the diagnosis of cancer via MRI images. A system to
automatically navigate a vehicle across Martian terrain. A system to inspect welds in an
automotive assembly factory.
- The image analysis process requires the use of tools such as image segmentation, image
transforms, feature extraction and pattern classification. Image segmentation is often one of the
first steps in finding higher level objects from the raw image data. Feature extraction is the
process of acquiring higher level image information, such as shape or color information, and may
require the use of image transforms to find spatial frequency information. Pattern classification
is the act of taking this higher level information and identifying objects within the image.
- hardware and software.
- Gigabyte Ethernet, USB 3.2, USB 4.0, Camera Link.
- It samples an analog video signal to create a digital image. This sampling is done at a fixed
rate when it measures the voltage of the signal and uses this value for the pixel brightness. It uses
the horizontal synch pulse to control timing for one line of video (one row in the digital image),
and the vertical synch pulse to tell the end of a field or frame
- A sensor is a measuring device that responds to various parts of the EM spectrum, or other
signal that we desire to measure. To create images the measurements are taken across a two-
dimensional gird, thus creating a digital image.
- A range image is an image where the pixel values correspond to the distance from the imaging
sensor. They are typically created with radar, ultrasound or lasers.
- The reflectance function describes the way an object reflects incident light. This relates to
what we call color and texture it determines how the object looks.
- Radiance is the light energy reflected from, or emitted by, an object; whereas irradiance is
the incident light falling on a surface. So radiance is measured in Power/(Area)(SolidAngle), and
irradiance is measure in Power./Area.
- A photon is a massless particle that is used to model EM radiation. A CCD is a charge-
coupled device. Quantum efficiency is a measure of how effectively a sensing element converts
photonic energy into electrical energy, and is given by the ratio of electrical output to photonic
input.
sec
- Yes,a2solida2statea2becausea2thea2quantuma2efficiencya2isa295%
Na2 a2 A ta2 a 2 b () q () d
a2 20 ( 10 a 2 x
3
700
700
)a 2
a 2
600 ( 0. 95 ) d
400
a 2
2 a
700
a2 114 a
a
d a2a2 114
a2 114 ( 165 , 000 )a2a2 1. 881x
7
400
a 2
2 a 2
400
- Witha2interlaceda2scanning,a2aa2framea2ina21/30a2ofa2aa2seconda2isa2aa2fielda2ratea2ofa21/60a2ofa
2 aa2second.a2Herea2wea2 240 a2linesa2pera2fielda2anda2 640 a2pixelsa2pera2linea2whicha2givesa2(240)(
0)a2=a2153,600a2pixelsa2ina21/60a2ofa2aa2second.a2Soa2thea2samplinga2ratea2musta2be:
153 , 600 a2pixelsa
a2 9. 216 a2x
6
a2 ;a2ora2abouta2 9 a2megahertz
- Gammaa2raysa2havea2thea2mosta2energy,a2radioa2wavesa2havea2thea2least.a2Fora2humana2life,a
morea2energya2isa2morea2dangerous.
- UVa2isa2useda2ina2fluorescencea2microscopy,a2anda2IRa2imagesa2area2useda2ina2remotea
sensing,a2lawa2enforcement,a2medicala2thermographya2anda2firea2detection.
- Acoustica2imaginga2worksa2bya2sendinga2outa2pulsesa2ofa2sonica2energya2(sound)a2ata2various
a2frequencies,a2anda2thena2measuringa2thea2reflecteda2waves.a2Thea2timea2ita2takesa2fora2thea2refle
cteda2signala2toa2appeara2containsa2distancea2information,a2anda2thea2amounta2ofa2energya2reflect
eda2containsa2informationa2abouta2thea2object’sa2densitya2anda2material.a2Thea2measureda2informa
tiona2isa2thena2useda2toa2createa2aa2twoa2ora2threea2dimensionala2image.a2Ita2isa2useda2ina2geologic
ala2applications,a2fora2examplea2oila2anda2minerala2exploration,a2typicallya2usea2lowa2frequencya
soundsa2(arounda2hundredsa2ofa2hertz).a2Ultrasonic,a2ora2higha2frequencya2sound,a2imaginga2isa2o
ftena2useda2ina2manufacturinga2toa2detecta2defectsa2anda2ina2medicinea2toa2“see”a2insidea2opaquea
2 objectsa2sucha2asa2aa2woman’sa2womba2toa2imagea2aa2developinga2baby.
- Electrona2microscopesa2cana2magnifya2mucha2morea2thana2lighta2microscopes,a21,000a2co
mpareda2toa2200,000a2times.a2Ita2usesa2aa2focuseda2beama2ofa2electronsa2insteada2ofa2lighta2ener
gya2toa2imagea2thea2objects.
- 1)a2structureda2lightinga2createda2witha2aa2lasera2anda2rotatinga2mirrors,a22)a2time-of-
flighta2usinga2aa2transmittera2anda2receivera2anda2variousa2typesa2ofa2signals.
- Ana2 opticala2imagea2 isa2aa2collectiona2ofa2spatiallya2distributeda2lighta2energy.a2Opticala2im
agesa2cana2bea2representeda2asa2videoa2informationa2ina2thea2forma2ofa2analoga2electricala2signal
s,a2anda2thesea2area2sampleda2toa2generatea2thea2digitala2imagea2 I(r,c).
- Aa2“real”a2imagea2isa2measureda2bya2aa2sensor,a2aa2computera2imagea2isa2generateda2bya2s
oftwarea2oftena2usinga2aa2mathematicala2model.
- Binarya2imagesa2area2 1 - bita2pera2pixela2(bpp),a2gray-
scalea2typicallya2 8 a2bpp,a2colora2typicallya2 24 a2bpp,a2anda2multispectrala2imagesa2cana2bea2manya
2 morea2bpp.a2Witha2fewera2bpp,a2wea2havea2lessa2information,a2buta2smallera2files.a2Binarya2imag
esa2containa2onlya2shapea2information,a2gray-
scalea2imagesa2containa2brightnessa2information,a2anda2colora2imagesa2havea2brightnessa2informa
tiona2typicallya2ina2threea2spectrala2bands.a2Multispectrala2imagesa2havea2morea2thana2 3 a2bandsa2a
nda2typicallya2includea2informationa2outsidea2ofa2thea2humana2visuala2spectrum.
- Oftena2toa2decouplea2thea2brightnessa2anda2thea2colora2information,a2whicha2createsa2aa2mor
ea2people-
a2orienteda2waya2ofa2describinga2colors.a2Hue/Saturation/Lightnessa2(HSL)a2colora2transforma2all
owsa2usa2toa2describea2colorsa2ina2termsa2thata2wea2cana2morea2readilya2understanda2(seea2Figure
a22.4-
3).a2Thea2 lightnessa2 (alsoa2referreda2asa2 intensitya2 ora2 value )a2isa2thea2brightnessa2ofa2thea2color,
a2anda2thea2 huea2 isa2whata2wea2normallya2thinka2ofa2asa2"color";a2fora2examplea2green,a2bluea2ora
2 orange.a2Thea2 saturationa2 isa2aa2measurea2ofa2howa2mucha2whitea2isa2ina2thea2color;a2fora2exam
ple,a2pinka2isa2reda2witha2morea2white,a2soa2ita2isa2lessa2saturateda2thana2aa2purea2red.
Supplementarya2Problems:
- Seea2figurea21.4-6.a2anda2applya2similara2triangles:
𝑏 ′a2 −a
=a 2 a
2
2 a 2
𝑐a 2 =a
- a)a
2 f
(𝑏 ′a2 −a2𝑏)a
𝑏a2−a
a2 numbera
fa
D
effective
.a2Thea2focala2length,a2 f ,a2isa2fixed.a2Wea2cana2changea2thea2f-numbera2by
changinga2thea2effectivea2diametera2bya2varyinga2thea2aperture.a2Asa2thea2effectivea2diametera2goe
sa2up,a2thea2f-
a2numbera2goesa2down,a2anda2vicea2versa.a 2 b)a2Thea2amounta2ofa2lighta2energya2thata2isa2intercep
teda2bya2thea2lensa2isa2propotionala2toa2thea2areaa2ona2whicha2thea2lighta2energya2falls.a2Ina2thisa2c
ase,a2thea2areaa2ofa2thea2circlea2ofa2thea2lensa2witha2diametera2D:a 2 Areaa2=a2πa2r
2
a2=a2π(D/2)
2
a 2 =
a2(π/4)D
2
a2.a2So,a2wea2cana2seea2thata2thea2areaa2isa2inversely
a
D
2
toa2thea2f-
numbera2squareda2= a
effectivea
2
.Sincea2thea2areaa2isa2propotionala2toa2thea2ligh
t
(a2 fa 2 a2 numbera 2 )
f
energy,a2ora2brightness,a 2 thea2imagea2brightnessa2isa2inverselya2proportionala2toa2thea2f-
numbera2squared.
b)a2 Alternatea2explanation :a2Measuringa2thea2brightnessa2bya2measuringa2thea2numbera2ofa2elect
ronsa2liberated,a2anda2usinga2 Ka2 asa2aa2constanta2toa2representa2thata2wea2havea2aa2fixeda2timea2int
ervala2anda2aa2constant
incidenta2photona2flux,a2wea2cana2say:
a
D
2
K
Na 2 a2 A ta2 a2 b () q () d a2a2 Brightnessa2 a2a2a2 a
2 a 2
eff
a2a2 a 2 Ka 2
D
2
eff
2 a 2
a
4( Brightnes
s )
K
a 2 a 2
2 a 2
D
eff
Solutionsa2fora2Chaptera23:a2Introductiona2toa2Digitala2Imagea2Analysis
- Imagea2analysisa2 involvesa2manipulatinga2thea2imagea2dataa2toa2determinea2exactlya2thea2in
formationa2requireda2toa2developa2thea2computera2imaginga2system.
Fora2computera2vision,a2thea2enda2producta2isa2typicallya2thea2extractiona2ofa2higha2levela2i
nformationa2fora2computera2analysisa2ora2manipulation.a2Thisa2higha2levela2informationa2maya2in
cludea2shapea2parametersa2toa2controla2aa2robotica2manipulator,a2terraina2analysisa2toa2enablea2aa
vehiclea2toa2navigatea2ona2mars,a2ora2colora2anda2texturea2featuresa2toa2helpa2ina2thea2diagnosisa
ofa2aa2skina2tumor.a2Imagea2analysisa2isa2centrala2toa2thea2computera2visiona2processa2anda2isa2oft
ena2uniquelya2associateda2witha2computera2vision;a2however,a2imagea2analysisa2isa2ana2important
a2toola2fora2imagea2processinga2applicationsa2asa2well.
Ina2imagea2processinga2fora2humana2visiona2applications,a2imagea2analysisa2methodsa2ma
ya2bea2useda2toa2helpa2determinea2thea2typea2ofa2processinga2requireda2anda2thea2specifica2parame
tersa2neededa2fora2thata2processing.a2Fora2example,a2developinga2ana2enhancementa2algorithma2(
Chaptera28),a2determininga2thea2degradationa2functiona2fora2ana2imagea2restorationa2procedurea2(
Chaptera29),a2anda2determininga2exactlya2whata2informationa2isa2visuallya2importanta2fora2ana2im
agea2compressiona2methoda2(Chaptera210)a2area2alla2imagea2analysisa2tasks.
- a)a21)a2Preprocessing,a22)a2Dataa2Reduction,a2anda23)a2Featurea2Analysis.a2b)a2S
patiala2anda2frequency/sequencya2domains.
- Figurea23.1-
3.a2 Preprocessing :a2Thea2preprocessinga2algorithms,a2techniquesa2anda2operatorsa2area2useda2toa
performa2initiala2processinga2thata2makesa2thea2primarya2dataa2reductiona2anda2analysisa2taska2easi
er.a2Theya2includea2operationsa2relateda2toa2extractinga2regionsa2ofa2interest,a2performinga2basica
mathematicala2operationsa2ona2images,a2simplea2enhancementa2ofa2specifica2imagea2features,a2an
da2dataa2reductiona2ina2botha2resolutiona2anda2brightness.a2Preprocessinga2isa2aa2stagea2wherea2the
a2requirementsa2area2typicallya2obviousa2anda2simple,a2sucha2asa2thea2removala2ofa2artifactsa2froma
2 images,a2ora2thea2eliminationa2ofa2imagea2information
thata2isa2nota2requireda2fora2thea2application.a2Aftera2preprocessinga2wea2cana2performa2 segmenta
tiona2 ona2thea2imagea2ina2thea2spatiala2domaina2ora2converta2ita2intoa2thea2frequencya2domaina2via
a2aa2mathematicala2 transform .a2Notea2thea2dotteda2linea2betweena2segmentationa2anda2thea2transfo
rma2block,a2thisa2isa2fora2extractinga2spectrala2featuresa2ona2segmenteda2partsa2ofa2thea2image.a2A
ftera2eithera2ofa2thesea2processesa2wea2maya2choosea2toa2 filtera2 thea2image.a2Thisa2filteringa2proce
ssa2furthera2reducesa2thea2dataa2anda2allowsa2usa2toa2 extracta2thea2featuresa2 thata2wea2maya2requir
ea2fora2 analysis .a2Aftera2thea2analysis,a2wea2havea2aa2 feedbacka2loopa2 thata2providesa2fora2ana2appl
ication-specifica2reviewa2ofa2thea2analysisa2results.
- Thea2imagea2geometrya2operationsa2includea2crop,a2zoom,a2enlarge,a2shrink,a2translate,a2anda2ro
tate.a2Cropa2cutsa2outa2aa2portiona2ofa2ana2image,a2zooma2enlargesa2aa2portiona2ofa2ana2image.a2En
largea2makesa2ana2entirea2imagea2bigger,a2whilea2shrinka2reducesa2thea2sizea2ofa2ana2image.a2Tran
slatea2movesa2ana2imagea2upa2ora2down,a2anda2rotatea2spinsa2ana2imagea2abouta2thea2imagea2origi
n,a2whicha2isa2typicallya2ina2thea2uppera2lefta2ora2center.
5. 3 a
6. 6 a
- a)a2(102,67)
r
=a2r(a2 cosa2 )+c(a 2 sina 2 a2 )a2 a2 42 (cosa 2 50 )a2a2 100 (sina2 50 )a2a2 104
b)
c ˆ =a2-a2r(a2 sina2a2 )+c(a2 cosa2 )a2 a2 42 (sina2 50 )a2a2 100 (cosa 2 50 )a2a2 32
a
- Meana2→a 2
1 a
a
Enhancementa2→a
a
0 a
2
0 a
a2 2 1
a2 2 1
a2 2
a2 2
- a)a2Ita2looksa2mostlya2graya2becausea2ina2thea2differencea2imagea2thea2numbersa2area2centereda
2 arounda2zeroa2–
a2negativea2anda2positivea2numbers.a2Duringa2thea2remapa2processa2thea2centera2value,a2ina2thisa2c
asea2zero,a2isa2shifteda2toa2thea2middlea2(128),a2whicha2createsa2aa2mostlya2graya2image.a2b)a2Toa
createa2thea2imagea2asa2showna2ina2Figa23.2-
7c,a2starta2witha2thea2subtracteda2imagea2createda2froma2(scenea2ba2–
a2scenea2a)a2anda2shifta2alla2valuesa2belowa2 128 a2toa2zeroa2anda2thena2stretcha2alla2thea2positivea2v
aluesa2toa2255,a2creatinga2 image1 .a2Thisa2cana2bea2donea2witha2aa2histograma2stretcha2anda2clipa2a
ta2thea2lowa2enda2ofa20.5a2ora250%.a2Next,a2histograma2stretcha2thea2originala2subtracteda2imagea
anda2clipa250%a2ata2higha2enda2anda2performa2aa2logicala2NOTa2ona2thea2result,a2creatinga2 image
2 .a2Nowa2wea2adda2 image1a2 anda2 image2 ,a2anda2thena2stretcha2thea2histograma2witha21%a2clippin
ga2ata2botha2ends.a2 Note:a2mosta2studentsa2willa2havea2finda2parta2(b)a2difficulta2untila2aftera2Cha
ptera28.
- Multiplicationa2anda2divisiona2cana2bea2useda2to:a21)a2darkena2ana2image,a22)a2brightena
ana2image,a23)a2speciala2effects,a24)a2adda2texturea2toa2aa2computera2generateda2image
2 a2;a
mapsa2toa2lowa2enda2ofa2range.
- Falsea2contouringa2refersa2toa2falsea2edges,a2ora2lines,a2thata2appeara2ina2thea2imagea2asa2aa2res
ulta2ofa2thea2graya2levela2quantization.a2Thisa2effecta2cana2bea2visuallya2improveda2bya2usinga2ana
2 IGSa2(improveda2graya2scale)a2quantizationa2method.a2Thea2IGSa2methoda2takesa2advantagea2ofa
2 thea2humana2visuala2system'sa2sensitivitya2toa2edgesa2bya2addinga2aa2smalla2randoma2numbera2to
a2eacha2pixela2beforea2quantization
- Variablea2bina2widtha2quantizationa2allowsa2fora2differenta2quantizationa2stepa2sizesa2toa2bea
useda2withina2aa2singlea2quantizationa2scheme.a2Ita2isa2morea2complexa2anda2nota2asa2fasta2toa2per
forma2asa2uniforma2bina2widtha2quantization.a2Ita2isa2typicallya2useda2fora2application-
specifica2reasonsa2wherea2thea2numbera2ofa2graya2levelsa2useda2toa2representa2thea2imagea2isa2sma
ll,a2anda2specifica2structuresa2area2beinga2analyzed.
- 1)a2averaging,a22)a2median,a2ora23)a2decimation.a2Fora2thea2firsta2method,a2averaging,a2wea2ta
kea2alla2thea2pixelsa2ina2eacha2groupa2anda2finda2thea2averagea2graya2levela2bya2summinga2thea2val
uesa2anda2dividinga2bya2thea2numbera2ofa2pixelsa2ina2thea2group.a2Witha2thea2seconda2method,a2me
dian,a2wea2sorta2alla2thea2pixela2valuesa2froma2lowesta2toa2highesta2anda2thena2selecta2thea2middlea
2 value.a2Thea2thirda2approach,a2decimation,a2alsoa2knowna2asa2sub-
sampling,a2entailsa2simplya2eliminatinga2somea2ofa2thea2data.a2Decimationa2isa2fastest,a2mediana2is
a2thea2slowest.
- Thea2 histograma2 ofa2ana2imagea2isa2aa2plota2ofa2graya2levela2versusa2thea2numbera2ofa2pixelsa
ina2thea2imagea2ata2eacha2graya2level.a2Ita2isa2usefula2toa2helpa2usa2finda2aa2thresholda2toa2separat
ea2objecta2anda2backgrounda2ina2ana2image.
- CVIPtools.a2Aa2gooda2thresholda2cannota2bea2founda2fora2thisa2imagea2duea2toa2thea2histo
grama2–a2graya2levelsa2ina2thea2objecta2area2thea2samea2asa2backgrounda2graya2levels.
a 2
B
1
O
1
B
2
a
4 - connectivity,a2Ba2fora2Background,a2Oa2fora2Object→a 2
O B Oa 2
a 2
Wea2havea2foura2separate
a 2 a 2
2 3 3 a
B
4
O
4
B
5 a
objects,a2soa2thea2backgrounda2shoulda2bea2connected,a2buta2ita2isa2not
i
i
B
1
O
1
B
1
a
8 - connectivity,a2Ba2fora2Background,a2Oa2fora2Object→a 2
O B Oa 2
a 2
Wea2havea2onea2connected
a 2 a 2
1 1 1 a
B
1
O
1
B
1 a
object,a2soa2thea2backgrounda2shoulda2nota2bea2connected,a2buta2ita2is
Threea2waysa2toa2avoida2this:a21)a2Usea2eight-connectivitya2fora2backgrounda2anda2four-
connectivitya2fora2thea2objects.a22)a2Usea2four-connectivitya2fora2backgrounda2anda2eight-
connectivitya2fora2thea2objects.
3)a2Usea2six-connectivity
- Thea2UPDATEa2blocka2isa2toa2deala2witha2thea2situationa2whena2aa2connecteda2objecta2isa2f
ounda2toa2havea2twoa2differenta2labels,a2asa2showna2ina2Fig.a23.3-5.
- Ofa2thea2featuresa2discusseda2ina2thisa2chaptera2thea2centera2ofa2areaa2(tellsa2usa2wherea2ita2i
s)a2anda2axisa2ofa2leasta2seconda2momenta2(tellsa2usa2howa2ita2isa2oriented/rotated)
- h(r)a2=a2[3,a23,a22];a 2 v(c)a2=a2[0,a21,a22,a23,a22]
- No,a2thea2numbera2ofa2objectsa2isa2nota2necessarilya2equala2toa2thea2numbera2ofa2convexit
ies.a2No,a2thea2numbera2ofa2holesa2isa2nota2necessarilya2equala2toa2thea2numbera2ofa2concavi
ties.a2Justa2because
Aa2 a2 Ba2 a2 Ca2 a2 Da2 ,a2doesa2nota2meana2A=Ca2anda2B=D.a2Ina2thisa2casea2wea2cana2havea2an
a2imagea2witha2manya2convexitiesa2anda2concavities,a2buta2onlya2onea2object,a2ifa2thea2objecta
hasa2aa2convoluteda2(curvy,a2wavy)a2border.a2(thea2convexitiesa2anda2concavitiesa2willa2cance
la2eacha2othera2out)
Supplementarya2Problems
- a)a2Areaa2=a2 16 a2(suma2alla2thea21’s);
ra 2 =a
2
a2 1
A
N-
1 a 2 N-
1
ra Ia 2 ia
r,c)
a
a2 1 a
a
1 a
a 21 a2a 21 a
a2 1
a 21 a 2 a2 1 a 2 a
22 a2
2 a2a2 3 a2a
23 a2
3 a2a2 3
a2
4 a 2 a
a 226
a
ca 2 =a
2
a2 1
A
N- 1 a 2 N-
1
ca Ia ia
(r,c)
a
a
1 a
2 a
1 a
a2 1 a2 2 a2a
2 a2
2 a2a2 3 a2 3 a2a
23 a2
3 a2a
23 a2
a2 4 a2a2 4 a
4 a2a2 4
ia 2 r=
a 2 c=
ia 2 r=
a 2 c=
b)
Na2- 1 a2Na2- 1
a 2 rcIa ia
( r , c )
tan (2a
)a2=a2 2 a
r=0a2c=
a
a2 2
a 2 a 2 1 a2 3 a2a2 0 a2 1 a2 6 a2a2 0 a2a2 3 a2a2 6 a2a2 9 a2 12 a2a2 0 a2a2 4 a2a2 8 a2 12 a2a2 5 a2a2 6 a 2 a
2
i
Na2-
1 a2Na2- 1
Na2- 1 a2Na2- 1
2 2
a 2 ra
2 a
I (r,c)a2 a 2
a 2 ca
2
2 a
I
(r,c)
( 6 ) 1 a 2 a2 2 ( 2 a 2 )a 2 a2 4 ( 9 )a 2 1 ( 16 ) a
a2 2 a2 12 a2a2 45 a2a
264 a2a2 25 a2a2 36
152
66 a2 18
4
i
r=0a2c=
a2 1. 2881356
i
r=0a2c=
a
a 2 a2 26. 1
∘
c) Note:a2ifa2therea2isa2aa2‘1’a2ona2ana2imagea2edge,a2thea2imagea2musta2bea2zero-padded!
Eulera2numbera2=a2(#a2convexities)a2–a2(#a2concavities)a2=a2 3 a2–a2 2 a2=a2 1
] [ ]
ora2,a2Eulera2numbera2=a2(#a2objects)a2–a2(#a2holes)a2=a2 1 a2–a2 0 a2=a2 1
d) Note:a2ifa2therea2isa2aa2‘1’a2ona2ana2imagea2edge,a2thea2imagea2musta2bea2zero-padded!
Eulera2numbera2=a2(#a2convexities)a2–a2(#a2concavities)a2=a2 2 a2–a2 1 a2=a2 1
] [ ]
ora2,a2Eulera2numbera2=a2(#a2objects)a2–a2(#a2holes)a2=a2 1 a2–a2 0 a2=a2 1
- CVIPtools.a2Thea2resultsa2dependa2ona2thea2imagesa 2 created.a2Thea2successa2ratea2willa2decr
easea2asa2thea2imagea2isa2blurreda2anda2noisea2isa2added.a2Addinga2noisea2willa2probablya2makea
thea2successa2ratea2decreasea2morea2thana2blurringa2thea2images.a2Newa2algorithma2developmen
ta2willa2requirea2experimentationa2anda2creativity.
- Discussiona2willa2dependa2ona2researcha2results.a2Onea2methoda2isa2Otsua2methoda2froma2Chaptera24.
- Resultsa2willa2varya2dependinga2ona2thea2imagesa2anda2objectsa2selected.
- a)a2meana2valuea2fora2image:a 2 [ 4 ( 0 )a2+a2 25 ( 1 )a2+a2 5 ( 2 )a2+a2 30 ( 3 )]/64a2≈a21.
Iterationa21:
𝑚 1 a 2 =a 2
35 a 2
[
)a 2 +a 230 ( 3 )
]a 2 ≈a 2 2.
[
[
𝑚 2 a 2 =a 2
29 a 2
[
)a 2 +a 225 ( 1 )
]a 2 ≈a 2 0.
2.86a2+a20.
𝑛𝑒𝑤a 2
=a21.
𝑇𝑜𝑙𝑑a 2 −a2𝑇𝑛𝑒𝑤a 2 =a 2 1.95a 2 −a 2 1.86a 2 =a 2 0.09, 𝑛𝑜𝑡a 2 <a2𝑙𝑖𝑚𝑖𝑡
Iterationa22:
1
a 2 =a 2
35 a 2
[ 5 ( 2 )a 2 +a 230 ( 3 )]a 2 ≈a 2 2.
𝑚 2 a 2 =a 2
29 a 2
[
)a 2 +a 225 ( 1 )
]a 2 ≈a 2 0.
2.86a2+a20.
𝑛𝑒𝑤a 2
=a21.
𝑇𝑜𝑙𝑑a 2 −a2𝑇𝑛𝑒𝑤a 2 =a 2 1.86a 2 −a 2 1.86a 2 =a 2 0.0, 0.0a2<a2𝑙𝑖𝑚𝑖𝑡a 2 ∴a2𝐷𝑜𝑛𝑒!
Answera2=a21.86a2≈a2 2
b) weighteda2averagea2froma2twoa2histograma2peaks:a
[ 25
( 1
)
( 3
)]a 2
≈a22.
25+
Iterationa21:
1
a 2 =a 2
30 a 2
[ 30 ( 3 )]a 2 =a 2 3.
2
a 2 =a 2
34 a 2
[ 4 ( 0 )a 2 +a 225 ( 1 )a 2 +a 25 ( 2 )]a 2 ≈a 2 1.
3.0a2+a21.
𝑛𝑒𝑤a 2
≈a22.
𝑜𝑙𝑑
a 2 −a2𝑇 𝑛𝑒𝑤
a 2 =a 2 2.09a 2 −a 2 2.015a 2 =a 2 0.075, 𝑛𝑜𝑡a 2 <a2𝑙𝑖𝑚𝑖𝑡
Iterationa22:
1
a 2 =a 2
30 a 2
[ 30 ( 3 )]a 2 =a 2 3.
2
a 2 =a 2
34 a 2
[ 4 ( 0 )a 2 +a 225 ( 1 )a 2 +a 25 ( 2 )]a 2 ≈a 2 1.
