| |
| |
| Preface | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| Machines that See? | |
| |
| |
| |
| Application Problems | |
| |
| |
| |
| A Preview of the Digital Image | |
| |
| |
| |
| Image Database Query | |
| |
| |
| |
| Inspecting Crossbars for Holes | |
| |
| |
| |
| Examining the Inside of a Human Head | |
| |
| |
| |
| Processing Scanned Text Pages | |
| |
| |
| |
| Accounting for Snow Cover Using a Satellite Image | |
| |
| |
| |
| Understanding a Scene of Parts | |
| |
| |
| |
| Operations on Images | |
| |
| |
| |
| Changing Pixels in Small Neighborhoods | |
| |
| |
| |
| Enhancing an Entire Image | |
| |
| |
| |
| Combining Multiple Images | |
| |
| |
| |
| Computing Features from an Image | |
| |
| |
| |
| Extracting Non-iconic Representations | |
| |
| |
| |
| The Good, the Bad, and the Ugly | |
| |
| |
| |
| Use of Computers and Software | |
| |
| |
| |
| Related Areas | |
| |
| |
| |
| The Rest of the Book | |
| |
| |
| |
| References | |
| |
| |
| |
| Additional Exercises | |
| |
| |
| |
| Imaging and Image Representation | |
| |
| |
| |
| Sensing Light | |
| |
| |
| |
| Imaging Devices | |
| |
| |
| |
| CCD Cameras | |
| |
| |
| |
| Image Formation | |
| |
| |
| |
| Video Cameras | |
| |
| |
| |
| The Human Eye | |
| |
| |
| |
| Problems in Digital Images* | |
| |
| |
| |
| Geometric Distortion | |
| |
| |
| |
| Scattering | |
| |
| |
| |
| Blooming | |
| |
| |
| |
| CCD Variations | |
| |
| |
| |
| Clipping or Wrap-Around | |
| |
| |
| |
| Chromatic Distortion | |
| |
| |
| |
| Quantization Effects | |
| |
| |
| |
| Picture Functions and Digital Images | |
| |
| |
| |
| Types of Images | |
| |
| |
| |
| Image Quantization and Spatial Measurement | |
| |
| |
| |
| Digital Image Formats* | |
| |
| |
| |
| Image File Header | |
| |
| |
| |
| Image Data | |
| |
| |
| |
| Data Compression | |
| |
| |
| |
| Commonly Used Formats | |
| |
| |
| |
| Run-Coded Binary Images | |
| |
| |
| |
| PGM: Portable Gray Map | |
| |
| |
| |
| GIF Image File Format | |
| |
| |
| |
| TIFF Image File Format | |
| |
| |
| |
| JPEG Format for Still Photos | |
| |
| |
| |
| PostScript | |
| |
| |
| |
| MPEG Format for Video | |
| |
| |
| |
| Comparison of Formats | |
| |
| |
| |
| Richness and Problems of Real Imagery | |
| |
| |
| |
| 3D Structure from 2D Images | |
| |
| |
| |
| Five Frames of Reference | |
| |
| |
| |
| Pixel Coordinate Frame I | |
| |
| |
| |
| Object Coordinate Frame O | |
| |
| |
| |
| Camera Coordinate Frame C | |
| |
| |
| |
| Real Image Coordinate Frame F | |
| |
| |
| |
| World Coordinate Frame W | |
| |
| |
| |
| Other Types of Sensors* | |
| |
| |
| |
| Microdensitometer* | |
| |
| |
| |
| Color and Multispectral Images* | |
| |
| |
| |
| X-ray* | |
| |
| |
| |
| Magnetic Resonance Imaging (MRI)* | |
| |
| |
| |
| Range Scanners and Range Images* | |
| |
| |
| |
| References | |
| |
| |
| |
| Binary Image Analysis | |
| |
| |
| |
| Pixels and Neighborhoods | |
| |
| |
| |
| Applying Masks to Images | |
| |
| |
| |
| Counting the Objects in an Image | |
| |
| |
| |
| Connected Components Labeling | |
| |
| |
| |
| Binary Image Morphology | |
| |
| |
| |
| Structuring Elements | |
| |
| |
| |
| Basic Operations | |
| |
| |
| |
| Some Applications of Binary Morphology | |
| |
| |
| |
| Conditional Dilation | |
| |
| |
| |
| Region Properties | |
| |
| |
| |
| Region Adjacency Graphs | |
| |
| |
| |
| Thresholding Gray-Scale Images | |
| |
| |
| |
| The Use of Histograms for Threshold Selection | |
| |
| |
| |
| Automatic Thresholding: The Otsu Method* | |
| |
| |
| |
| References | |
| |
| |
| |
| Pattern Recognition Concepts | |
| |
| |
| |
| Pattern Recognition Problems | |
| |
| |
| |
| Common Model for Classification | |
| |
| |
| |
| Classes | |
| |
| |
| |
| Sensor/Transducer | |
| |
| |
| |
| Feature Extractor | |
| |
| |
| |
| Classifier | |
| |
| |
| |
| Building the Classification System | |
| |
| |
| |
| Evaluation of System Error | |
| |
| |
| |
| False Alarms and False Dismissals | |
| |
| |
| |
| Precision Versus Recall | |
| |
| |
| |
| Features Used for Representation | |
| |
| |
| |
| Feature Vector Representation | |
| |
| |
| |
| Implementing the Classifier | |
| |
| |
| |
| Classification Using the Nearest Class Mean | |
| |
| |
| |
| Classification Using the Nearest Neighbors | |
| |
| |
| |
| Structural Techniques | |
| |
| |
| |
| The Confusion Matrix | |
| |
| |
| |
| Decision Trees | |
| |
| |
| |
| Bayesian Decision-Making | |
| |
| |
| |
| Parametric Models for Distributions | |
| |
| |
| |
| Decisions Using Multidimensional Data | |
| |
| |
| |
| Machines that Learn | |
| |
| |
| |
| Artificial Neural Nets* | |
| |
| |
| |
| The Perceptron Model | |
| |
| |
| |
| The Multilayer Feedforward Network | |
| |
| |
| |
| References | |
| |
| |
| |
| Filtering and Enhancing Images | |
| |
| |
| |
| What Needs Fixing? | |
| |
| |
| |
| An Image Needs Improvement | |
| |
| |
| |
| Low-Level Features Must Be Detected | |
| |
| |
| |
| Gray-Level Mapping | |
| |
| |
| |
| Histogram Equalization | |
| |
| |
| |
| Removal of Small Image Regions | |
| |
| |
| |
| Removal of Salt-and-Pepper Noise | |
| |
| |
| |
| Removal of Small Components | |
| |
| |
| |
| Image Smoothing | |
| |
| |
| |
| Median Filtering | |
| |
| |
| |
| Computing an Output Image from an Input Image | |
| |
| |
| |
| Detecting Edges Using Differencing Masks | |
| |
| |
| |
| Differencing 1D Signals | |
| |
| |
| |
| Difference Operators for 2D Images | |
| |
| |
| |
| Gaussian Filtering and LOG Edge Detection | |
| |
| |
| |
| Detecting Edges with the LOG Filter | |
| |
| |
| |
| On Human Edge Detection | |
| |
| |
| |
| Marr-Hildreth Theory | |
| |
| |
| |
| The Canny Edge Detector | |
| |
| |
| |
| Masks as Matched Filters* | |
| |
| |
| |
| The Vector Space of All Signals of n Samples | |
| |
| |
| |
| Using an Orthogonal Basis | |
| |
| |
| |
| Cauchy-Schwartz Inequality | |
| |
| |
| |
| The Vector Space of m [times] n Images | |
| |
| |
| |
| A Roberts Basis for 2 [times] 2 Neighborhoods | |
| |
| |
| |
| The Frei-Chen Basis for 3 [times] 3 Neighborhoods | |
| |
| |
| |
| Convolution and Cross Correlation* | |
| |
| |
| |
| Defining Operations via Masks | |
| |
| |
| |
| The Convolution Operation | |
| |
| |
| |
| Possible Parallel Implementations | |
| |
| |
| |
| Analysis of Spatial Frequency using Sinusoids* | |
| |
| |
| |
| A Fourier Basis | |
| |
| |
| |
| 2D Picture Functions | |
| |
| |
| |
| Discrete Fourier Transform | |
| |
| |
| |
| Bandpass Filtering | |
| |
| |
| |
| Discussion of the Fourier Transform | |
| |
| |
| |
| The Convolution Theorem* | |
| |
| |
| |
| Summary and Discussion | |
| |
| |
| |
| References | |
| |
| |
| |
| Color and Shading | |
| |
| |
| |
| Some Physics of Color | |
| |
| |
| |
| Sensing Illuminated Objects | |
| |
| |
| |
| Additional Factors | |
| |
| |
| |
| Sensitivity of Receptors | |
| |
| |
| |
| The RGB Basis for Color | |
| |
| |
| |
| Other Color Bases | |
| |
| |
| |
| The CMY Subtractive Color System | |
| |
| |
| |
| HSI: Hue-Saturation-Intensity | |
| |
| |
| |
| YIQ and YUV for TV Signals | |
| |
| |
| |
| Using Color for Classification | |
| |
| |
| |
| Color Histograms | |
| |
| |
| |
| Color Segmentation | |
| |
| |
| |
| Shading | |
| |
| |
| |
| Radiation from One Light Source | |
| |
| |
| |
| Diffuse Reflection | |
| |
| |
| |
| Specular Reflection | |
| |
| |
| |
| Darkening with Distance | |
| |
| |
| |
| Complications | |
| |
| |
| |
| Phong Model of Shading* | |
| |
| |
| |
| Human Perception Using Shading | |
| |
| |
| |
| Related Topics* | |
| |
| |
| |
| Applications | |
| |
| |
| |
| Human Color Perception | |
| |
| |
| |
| Multispectral Images | |
| |
| |
| |
| Thematic Images | |
| |
| |
| |
| References | |
| |
| |
| |
| Texture | |
| |
| |
| |
| Texture, Texels, and Statistics | |
| |
| |
| |
| Texel-Based Texture Descriptions | |
| |
| |
| |
| Quantitative Texture Measures | |
| |
| |
| |
| Edge Density and Direction | |
| |
| |
| |
| Local Binary Partition | |
| |
| |
| |
| Co-occurrence Matrices and Features | |
| |
| |
| |
| Laws Texture Energy Measures | |
| |
| |
| |
| Autocorrelation and Power Spectrum | |
| |
| |
| |
| Texture Segmentation | |
| |
| |
| |
| References | |
| |
| |
| |
| Content-Based Image Retrieval | |
| |
| |
| |
| Image Database Examples | |
| |
| |
| |
| Image Database Queries | |
| |
| |
| |
| Query-by-Example | |
| |
| |
| |
| Image Distance Measures | |
| |
| |
| |
| Color Similarity Measures | |
| |
| |
| |
| Texture Similarity Measures | |
| |
| |
| |
| Shape Similarity Measures | |
| |
| |
| |
| Object Presence and Relational Similarity Measures | |
| |
| |
| |
| Database Organization | |
| |
| |
| |
| Standard Indexes | |
| |
| |
| |
| Spatial Indexing | |
| |
| |
| |
| Indexing for Content-Based Image Retrieval with Multiple Distance Measures | |
| |
| |
| |
| References | |
| |
| |
| |
| Motion From 2D Image Sequences | |
| |
| |
| |
| Motion Phenomena and Applications | |
| |
| |
| |
| Image Subtraction | |
| |
| |
| |
| Computing Motion Vectors | |
| |
| |
| |
| The Decathlete Game | |
| |
| |
| |
| Using Point Correspondences | |
| |
| |
| |
| MPEG Compression of Video | |
| |
| |
| |
| Computing Image Flow* | |
| |
| |
| |
| The Image Flow Equation* | |
| |
| |
| |
| Solving for Image Flow by Propagating Constraints* | |
| |
| |
| |
| Computing the Paths of Moving Points | |
| |
| |
| |
| Integrated Problem-Specific Tracking | |
| |
| |
| |
| Detecting Significant Changes in Video | |
| |
| |
| |
| Segmenting Video Sequences | |
| |
| |
| |
| Ignoring Certain Camera Effects | |
| |
| |
| |
| Storing Video Subsequences | |
| |
| |
| |
| References | |
| |
| |
| |
| Image Segmentation | |
| |
| |
| |
| Identifying Regions | |
| |
| |
| |
| Clustering Methods | |
| |
| |
| |
| Region Growing | |
| |
| |
| |
| Representing Regions | |
| |
| |
| |
| Overlays | |
| |
| |
| |
| Labeled Images | |
| |
| |
| |
| Boundary Coding | |
| |
| |
| |
| Quadtrees | |
| |
| |
| |
| Property Tables | |
| |
| |
| |
| Identifying Contours | |
| |
| |
| |
| Tracking Existing Region Boundaries | |
| |
| |
| |
| The Canny Edge Detector and Linker | |
| |
| |
| |
| Aggregating Consistent Neighboring Edgels into Curves | |
| |
| |
| |
| Hough Transform for Lines and Circular Arcs | |
| |
| |
| |
| Fitting Models to Segments | |
| |
| |
| |
| Identifying Higher-level Structure | |
| |
| |
| |
| Ribbons | |
| |
| |
| |
| Detecting Corners | |
| |
| |
| |
| Segmentation Using Motion Coherence | |
| |
| |
| |
| Boundaries in Space-Time | |
| |
| |
| |
| Aggregrating Motion Trajectories | |
| |
| |
| |
| References | |
| |
| |
| |
| Matching In 2D | |
| |
| |
| |
| Registration of 2D Data | |
| |
| |
| |
| Representation of Points | |
| |
| |
| |
| Affine Mapping Functions | |
| |
| |
| |
| A Best 2D Affine Transformation* | |
| |
| |
| |
| 2D Object Recognition via Affine Mapping | |
| |
| |
| |
| 2D Object Recognition via Relational Matching | |
| |
| |
| |
| Nonlinear Warping | |
| |
| |
| |
| Summary | |
| |
| |
| |
| References | |
| |
| |
| |
| Perceiving 3D From 2D Images | |
| |
| |
| |
| Intrinsic Images | |
| |
| |
| |
| Labeling of Line Drawings from Blocks World | |
| |
| |
| |
| 3D Cues Available in 2D Images | |
| |
| |
| |
| Other Phenomena | |
| |
| |
| |
| Shape from X | |
| |
| |
| |
| Vanishing Points | |
| |
| |
| |
| Depth from Focus | |
| |
| |
| |
| Motion Phenomena | |
| |
| |
| |
| Boundaries and Virtual Lines | |
| |
| |
| |
| Alignments are Non-accidental | |
| |
| |
| |
| The Perspective Imaging Model | |
| |
| |
| |
| Depth Perception from Stereo | |
| |
| |
| |
| Establishing Correspondences | |
| |
| |
| |
| The Thin Lens Equation* | |
| |
| |
| |
| Concluding Discussion | |
| |
| |
| |
| References | |
| |
| |
| |
| 3D Sensing and Object Pose Computation | |
| |
| |
| |
| General Stereo Configuration | |
| |
| |
| |
| 3D Affine Transformations | |
| |
| |
| |
| Coordinate Frames | |
| |
| |
| |
| Translation | |
| |
| |
| |
| Scaling | |
| |
| |
| |
| Rotation | |
| |
| |
| |
| Arbitrary Rotation | |
| |
| |
| |
| Alignment via Transformation Calculus | |
| |
| |
| |
| Camera Model | |
| |
| |
| |
| Perspective Transformation Matrix | |
| |
| |
| |
| Orthographic and Weak Perspective Projections | |
| |
| |
| |
| Computing 3D Points Using Multiple Cameras | |
| |
| |
| |
| Best Affine Calibration Matrix | |
| |
| |
| |
| Calibration Jig | |
| |
| |
| |
| Defining the Least-Squares Problem | |
| |
| |
| |
| Discussion of the Affine Method | |
| |
| |
| |
| Using Structured Light | |
| |
| |
| |
| A Simple Pose Estimation Procedure | |
| |
| |
| |
| An Improved Camera Calibration Method* | |
| |
| |
| |
| Intrinsic Camera Parameters | |
| |
| |
| |
| Extrinsic Camera Parameters | |
| |
| |
| |
| Calibration Example | |
| |
| |
| |
| Pose Estimation* | |
| |
| |
| |
| Pose from 2D-3D Point Correspondences | |
| |
| |
| |
| Constrained Linear Optimization | |
| |
| |
| |
| Computing the Transformation Tr = [R, T] | |
| |
| |
| |
| Verification and Optimization of Pose | |
| |
| |
| |
| 3D Object Reconstruction | |
| |
| |
| |
| Data Acquisition | |
| |
| |
| |
| Registration of Views | |
| |
| |
| |
| Surface Reconstruction | |
| |
| |
| |
| Space-Carving | |
| |
| |
| |
| Computing Shape from Shading | |
| |
| |
| |
| Photometric Stereo | |
| |
| |
| |
| Integrating Spatial Constraints | |
| |
| |
| |
| Structure from Motion | |
| |
| |
| |
| References | |
| |
| |
| |
| 3D Models and Matching | |
| |
| |
| |
| Survey of Common Representation Methods | |
| |
| |
| |
| 3D Mesh Models | |
| |
| |
| |
| Surface-Edge-Vertex Models | |
| |
| |
| |
| Generalized-Cylinder Models | |
| |
| |
| |
| Octrees | |
| |
| |
| |
| Superquadrics | |
| |
| |
| |
| True 3D Models versus View-Class Models | |
| |
| |
| |
| Physics-Based and Deformable Models | |
| |
| |
| |
| Snakes: Active Contour Models | |
| |
| |
| |
| Balloon Models for 3D | |
| |
| |
| |
| Modeling Motion of the Human Heart | |
| |
| |
| |
| 3D Object Recognition Paradigms | |
| |
| |
| |
| Matching Geometric Models via Alignment | |
| |
| |
| |
| Matching Relational Models | |
| |
| |
| |
| Matching Functional Models | |
| |
| |
| |
| Recognition by Appearance | |
| |
| |
| |
| References | |
| |
| |
| |
| Virtual Reality | |
| |
| |
| |
| Features of Virtual Reality Systems | |
| |
| |
| |
| Applications of VR | |
| |
| |
| |
| Augmented Reality (AR) | |
| |
| |
| |
| Teleoperation | |
| |
| |
| |
| Virtual Reality Devices | |
| |
| |
| |
| Summary of Sensing Devices for VR | |
| |
| |
| |
| Rendering Simple 3D Models | |
| |
| |
| |
| Composing Real and Synthetic Imagery | |
| |
| |
| |
| HCI and Psychological Issues | |
| |
| |
| |
| References | |
| |
| |
| |
| Case Studies | |
| |
| |
| |
| Veggie Vision: A System for Checking Out Vegetables | |
| |
| |
| |
| Application Domain and Requirements | |
| |
| |
| |
| System Design | |
| |
| |
| |
| Identification Procedure | |
| |
| |
| |
| More Details on the Process | |
| |
| |
| |
| Performance | |
| |
| |
| |
| Identifying Humans via the Iris of an Eye | |
| |
| |
| |
| Requirements for Identification Systems | |
| |
| |
| |
| System Design | |
| |
| |
| |
| Performance | |
| |
| |
| |
| References | |