Preface15
Acknowledgments19
TheBookWebSite20
AbouttheAuthors21
Chapter1Introduction23
1.1WhatIsDigitalImageProcessing?23
1.2TheOriginsofDigitalImageProcessing25
1.3ExamplesofFieldsthatUseDigitalImageProcessing29
1.3.1Gamma-RayImaging30
1.3.2X-RayImaging31
1.3.3ImagingintheUltravioletBand33
1.3.4ImagingintheVisibleandInfraredBands34
1.3.5ImagingintheMicrowaveBand40
1.3.6ImagingintheRadioBand42
1.3.7ExamplesinwhichOtherImagingModalitiesAreUsed42
1.4FundamentalStepsinDigitalImageProcessing47
1.5ComponentsofanImageProcessingSystem50
Summary53
ReferencesandFurtherReading53
Chapter2DigitalImageFundamentals57
2.1ElementsofVisualPerception58
2.1.1StructureoftheHumanEye58
2.1.2ImageFormationintheEye60
2.1.3BrightnessAdaptationandDiscrimination61
2.2LightandtheElectromagneticSpectrum65
2.3ImageSensingandAcquisition68
2.3.1ImageAcquisitionUsingaSingleSensor70
2.3.2ImageAcquisitionUsingSensorStrips70
2.3.3ImageAcquisitionUsingSensorArrays72
2.3.4ASimpleImageFormationModel72
2.4ImageSamplingandQuantization74
2.4.1BasicConceptsinSamplingandQuantization74
2.4.2RepresentingDigitalImages77
2.4.3SpatialandIntensityResolution81
2.4.4ImageInterpolation87
2.5SomeBasicRelationshipsbetweenPixels90
2.5.1NeighborsofaPixel90
2.5.2Adjacency,Connectivity,Regions,andBoundaries90
2.5.3DistanceMeasures93
2.6AnIntroductiontotheMathematicalToolsUsedinDigitalImageProcessing94
2.6.1ArrayversusMatrixOperations94
2.6.2LinearversusNonlinearOperations95
2.6.3ArithmeticOperations96
2.6.4SetandLogicalOperations102
2.6.5SpatialOperations107
2.6.6VectorandMatrixOperations114
2.6.7ImageTransforms115
2.6.8ProbabilisticMethods118
Summary120
ReferencesandFurtherReading120
Problems121
Chapter3IntensityTransformationsandSpatialFiltering126
3.1Background127
3.1.1TheBasicsofIntensityTransformationsandSpatialFiltering127
3.1.2AbouttheExamplesinThisChapter129
3.2SomeBasicIntensityTransformationFunctions129
3.2.1ImageNegatives130
3.2.2LogTransformations131
3.2.3Power-Law(Gamma)Transformations132
3.2.4Piecewise-LinearTransformationFunctions137
3.3HistogramProcessing142
3.3.1HistogramEqualization144
3.3.2HistogramMatching(Specification)150
3.3.3LocalHistogramProcessing161
3.3.4UsingHistogramStatisticsforImageEnhancement161
3.4FundamentalsofSpatialFiltering166
3.4.1TheMechanicsofSpatialFiltering167
3.4.2SpatialCorrelationandConvolution168
3.4.3VectorRepresentationofLinearFiltering172
3.4.4GeneratingSpatialFilterMasks173
3.5SmoothingSpatialFilters174
3.5.1SmoothingLinearFilters174
3.5.2Order-Statistic(Nonlinear)Filters178
3.6SharpeningSpatialFilters179
3.6.1Foundation180
3.6.2UsingtheSecondDerivativeforImageSharpening-TheLaplacian182
3.6.3UnsharpMaskingandHighboostFiltering184
3.6.4UsingFirst-OrderDerivativesfor(Nonlinear)ImageSharpening—TheGradient187
3.7CombiningSpatialEnhancementMethods191
3.8UsingFuzzyTechniquesforIntensityTransformationsandSpatialFiltering195
3.8.1Introduction195
3.8.2PrinciplesofFuzzySetTheory196
3.8.3UsingFuzzySets200
3.8.4UsingFuzzySetsforIntensityTransformations208
3.8.5UsingFuzzySetsforSpatialFiltering211
Summary214
ReferencesandFurtherReading214
Problems215
Chapter4FilteringintheFrequencyDomain221
4.1Background222
4.1.1ABriefHistoryoftheFourierSeriesandTransform222
4.1.2AbouttheExamplesinthisChapter223
4.2PreliminaryConcepts224
4.2.1ComplexNumbers224
4.2.2FourierSeries225
4.2.3ImpulsesandTheirSiftingProperty225
4.2.4TheFourierTransformofFunctionsofOneContinuousVariable227
4.2.5Convolution231
4.3SamplingandtheFourierTransformofSampledFunctions233
4.3.1Sampling233
4.3.2TheFourierTransformofSampledFunctions234
4.3.3TheSamplingTheorem235
4.3.4Aliasing239
4.3.5FunctionReconstruction(Recovery)fromSampledData241
4.4TheDiscreteFourierTransform(DFT)ofOneVariable242
4.4.1ObtainingtheDFTfromtheContinuousTransformofaSampledFunction243
4.4.2RelationshipBetweentheSamplingandFrequencyIntervals245
4.5ExtensiontoFunctionsofTwoVariables247
4.5.1The2-DImpulseandItsSiftingProperty247
4.5.2The2-DContinuousFourierTransformPair248
4.5.3Two-DimensionalSamplingandthe2-DSamplingTheorem249
4.5.4AliasinginImages250
4.5.5The2-DDiscreteFourierTransformandItsInverse257
4.6SomePropertiesofthe2-DDiscreteFourierTransform258
4.6.1RelationshipsBetweenSpatialandFrequencyIntervals258
4.6.2TranslationandRotation258
4.6.3Periodicity259
4.6.4SymmetryProperties261
4.6.5FourierSpectrumandPhaseAngle267
4.6.6The2-DConvolutionTheorem271
4.6.7Summaryof2-DDiscreteFourierTransformProperties275
4.7TheBasicsofFilteringintheFrequencyDomain277
4.7.1AdditionalCharacteristicsoftheFrequencyDomain277
4.7.2FrequencyDomainFilteringFundamentals279
4.7.3SummaryofStepsforFilteringintheFrequencyDomain285
4.7.4CorrespondenceBetweenFilteringintheSpatialandFrequencyDomains285
4.8ImageSmoothingUsingFrequencyDomainFilters291
4.8.1IdealLowpassFilters291
4.8.2ButterworthLowpassFilters295
4.8.3GaussianLowpassFilters298
4.8.4AdditionalExamplesofLowpassFiltering299
4.9ImageSharpeningUsingFrequencyDomainFilters302
4.9.1IdealHighpassFilters303
4.9.2ButterworthHighpassFilters306
4.9.3GaussianHighpassFilters307
4.9.4TheLaplacianintheFrequencyDomain308
4.9.5UnsharpMasking,HighboostFiltering,andHigh-Frequency-EmphasisFiltering310
4.9.6HomomorphicFiltering311
4.10SelectiveFiltering316
4.10.1BandrejectandBandpassFilters316
4.10.2NotchFilters316
4.11Implementation320
4.11.1Separabilityofthe2-DDFT320
4.11.2ComputingtheIDFTUsingaDFTAlgorithm321
4.11.3TheFastFourierTransform(FFT)321
4.11.4SomeCommentsonFilterDesign325
Summary325
ReferencesandFurtherReading326
Problems326
Chapter5ImageRestorationandReconstruction333
5.1AModeloftheImageDegradation/RestorationProcess334
5.2NoiseModels335
5.2.1SpatialandFrequencyPropertiesofNoise335
5.2.2SomeImportantNoiseProbabilityDensityFunctions336
5.2.3PeriodicNoise340
5.2.4EstimationofNoiseParameters341
5.3RestorationinthePresenceofNoiseOnly—SpatialFiltering344
5.3.1MeanFilters344
5.3.2Order-StatisticFilters347
5.3.3AdaptiveFilters352
5.4PeriodicNoiseReductionbyFrequencyDomainFiltering357
5.4.1BandrejectFilters357
5.4.2BandpassFilters358
5.4.3NotchFilters359
5.4.4OptimumNotchFiltering360
5.5Linear,Position-InvariantDegradations365
5.6EstimatingtheDegradationFunction368
5.6.1EstimationbyImageObservation368
5.6.2EstimationbyExperimentation369
5.6.3EstimationbyModeling369
5.7InverseFiltering373
5.8MinimumMeanSquareError(Wiener)Filtering374
5.9ConstrainedLeastSquaresFiltering379
5.10GeometricMeanFilter383
5.11ImageReconstructionfromProjections384
5.11.1Introduction384
5.11.2PrinciplesofComputedTomography(CT)387
5.11.3ProjectionsandtheRadonTransform390
5.11.4TheFourier-SliceTheorem396
5.11.5ReconstructionUsingParallel-BeamFilteredBackprojections397
5.11.6ReconstructionUsingFan-BeamFilteredBackprojections403
Summary409
ReferencesandFurtherReading410
Problems411
Chapter6ColorImageProcessing416
6.1ColorFundamentals417
6.2ColorModels423
6.2.1TheRGBColorModel424
6.2.2TheCMYandCMYKColorModels428
6.2.3TheHSIColorModel429
6.3PseudocolorImageProcessing436
6.3.1IntensitySlicing437
6.3.2IntensitytoColorTransformations440
6.4BasicsofFull-ColorImageProcessing446
6.5ColorTransformations448
6.5.1Formulation448
6.5.2ColorComplements452
6.5.3ColorSlicing453
6.5.4ToneandColorCorrections455
6.5.5HistogramProcessing460
6.6SmoothingandSharpening461
6.6.1ColorImageSmoothing461
6.6.2ColorImageSharpening464
6.7ImageSegmentationBasedonColor465
6.7.1SegmentationinHSIColorSpace465
6.7.2SegmentationinRGBVectorSpace467
6.7.3ColorEdgeDetection469
6.8NoiseinColorImages473
6.9ColorImageCompression476
Summary477
ReferencesandFurtherReading478
Problems478
Chapter7WaveletsandMultiresolutionProcessing483
7.1Background484
7.1.1ImagePyramids485
7.1.2SubbandCoding488
7.1.3TheHaarTransform496
7.2MultiresolutionExpansions499
7.2.1SeriesExpansions499
7.2.2ScalingFunctions501
7.2.3WaveletFunctions505
7.3WaveletTransformsinOneDimension508
7.3.1TheWaveletSeriesExpansions508
7.3.2TheDiscreteWaveletTransform510
7.3.3TheContinuousWaveletTransform513
7.4TheFastWaveletTransform515
7.5WaveletTransformsinTwoDimensions523
7.6WaveletPackets532
Summary542
ReferencesandFurtherReading542
Problems543
Chapter8ImageCompression547
8.1Fundamentals548
8.1.1CodingRedundancy550
8.1.2SpatialandTemporalRedundancy551
8.1.3IrrelevantInformation552
8.1.4MeasuringImageInformation553
8.1.5FidelityCriteria556
8.1.6ImageCompressionModels558
8.1.7ImageFormats,Containers,andCompressionStandards560
8.2SomeBasicCompressionMethods564
8.2.1HuffmanCoding564
8.2.2GolombCoding566
8.2.3ArithmeticCoding570
8.2.4LZWCoding573
8.2.5Run-LengthCoding575
8.2.6Symbol-BasedCoding581
8.2.7Bit-PlaneCoding584
8.2.8BlockTransformCoding588
8.2.9PredictiveCoding606
8.2.10WaveletCoding626
8.3DigitalImageWatermarking636
Summary643
ReferencesandFurtherReading644
Problems645
Chapter9MorphologicalImageProcessing649
9.1Preliminaries650
9.2ErosionandDilation652
9.2.1Erosion653
9.2.2Dilation655
9.2.3Duality657
9.3OpeningandClosing657
9.4TheHit-or-MissTransformation662
9.5SomeBasicMorphologicalAlgorithms664
9.5.1BoundaryExtraction664
9.5.2HoleFilling665
9.5.3ExtractionofConnectedComponents667
9.5.4ConvexHull669
9.5.5Thinning671
9.5.6Thickening672
9.5.7Skeletons673
9.5.8Pruning676
9.5.9MorphologicalReconstruction678
9.5.10SummaryofMorphologicalOperationsonBinaryImages684
9.6Gray-ScaleMorphology687
9.6.1ErosionandDilation688
9.6.2OpeningandClosing690
9.6.3SomeBasicGray-ScaleMorphologicalAlgorithms692
9.6.4Gray-ScaleMorphologicalReconstruction698
Summary701
ReferencesandFurtherReading701
Problems702
Chapter10ImageSegmentation711
10.1Fundamentals712
10.2Point,Line,andEdgeDetection714
10.2.1Background714
10.2.2DetectionofIsolatedPoints718
10.2.3LineDetection719
10.2.4EdgeModels722
10.2.5BasicEdgeDetection728
10.2.6MoreAdvancedTechniquesforEdgeDetection736
10.2.7EdgeLinkingandBoundaryDetection747
10.3Thresholding760
10.3.1Foundation760
10.3.2BasicGlobalThresholding763
10.3.3OptimumGlobalThresholdingUsingOtsu’sMethod764
10.3.4UsingImageSmoothingtoImproveGlobalThresholding769
10.3.5UsingEdgestoImproveGlobalThresholding771
10.3.6MultipleThresholds774
10.3.7VariableThresholding778
10.3.8MultivariableThresholding783
10.4Region-BasedSegmentation785
10.4.1RegionGrowing785
10.4.2RegionSplittingandMerging788
10.5SegmentationUsingMorphologicalWatersheds791
10.5.1Background791
10.5.2DamConstruction794
10.5.3WatershedSegmentationAlgorithm796
10.5.4TheUseofMarkers798
10.6TheUseofMotioninSegmentation800
10.6.1SpatialTechniques800
10.6.2FrequencyDomainTechniques804
Summary807
ReferencesandFurtherReading807
Problems809
Chapter11RepresentationandDescription817
11.1Representation818
11.1.1Boundary(Border)Following818
11.1.2ChainCodes820
11.1.3PolygonalApproximationsUsingMinimum-PerimeterPolygons823
11.1.4OtherPolygonalApproximationApproaches829
11.1.5Signatures830
11.1.6BoundarySegments832
11.1.7Skeletons834
11.2BoundaryDescriptors837
11.2.1SomeSimpleDescriptors837
11.2.2ShapeNumbers838
11.2.3FourierDescriptors840
11.2.4StatisticalMoments843
11.3RegionalDescriptors844
11.3.1SomeSimpleDescriptors844
11.3.2TopologicalDescriptors845
11.3.3Texture849
11.3.4MomentInvariants861
11.4UseofPrincipalComponentsforDescription864
11.5RelationalDescriptors874
Summary878
ReferencesandFurtherReading878
Problems879
Chapter12ObjectRecognition883
12.1PatternsandPatternClasses883
12.2RecognitionBasedonDecision-TheoreticMethods888
12.2.1Matching888
12.2.2OptimumStatisticalClassifiers894
12.2.3NeuralNetworks904
12.3StructuralMethods925
12.3.1MatchingShapeNumbers925
12.3.2StringMatching926
Summary928
ReferencesandFurtherReading928
Problems929
AppendixA932
Bibliography937
Index965