Image Analysis

Name: Image Analysis | 17th Scandinavian Conference, SCIA 2011, Ystad, Sweden, May 2011. Proceedings
Brand: Springer
Price: 128.39 EUR
Availability: OnlineOnly

17th Scandinavian Conference, SCIA 2011, Ystad, Sweden, May 2011. Proceedings

Anders Heyden Fredrik Kahl(Editor)

Springer (Publisher)

Published on 16. May 2011

XVII, 809 pages

E-Book

PDF with digital watermarking

System requirements

978-3-642-21227-7 (ISBN)

€128.39incl. 7% vat

System requirements

for PDF with digital watermarking

E-Book Single Licence

Available for download

Description

More details

Other editions

Content

Title Page
Preface
Organization
Table of Contents
Multiple View Geometry
Camera Self-calibration with Parallel Screw Axis Motion by Intersecting Imaged Horopters
Introduction
Background
Stratified Self-calibration
Horopter Curves
Problem Statement and Constraints
Self-Calibration Method
Affine Self-calibration by Intersecting the Imaged Horopters
Linear Upgrade from Affine to Euclidean
Experiments
Conclusion
References
Triangulating a Plane
Introduction
Quasiconvex Optimization
3D-Plane Triangulation
Homography from a Plane
Chirality
Experiments
Stability of the Proposed Formulation
Outlier Removal and Estimation
Conclusions
References
Human 3D Motion Computation from a Varying Number of Cameras
Introduction
Initial 3D Reconstruction
Imposing Weak Priors on the Initial Reconstruction
Link Length and Joint Angle Constraints
Temporal Smoothness, Link Length and Joint Angle Constraints
Results
Conclusion
References
Segmentation
Generalized Hard Constraints for Graph Segmentation
Introduction
Preliminaries
Edge Weighted Graphs
Graph Partitioning
Constrained Graph Cuts
Regional Constraints
Boundary Constraints
Strategies for Computing Constrained Graph Cuts
A General Strategy for Computing Constrained Graph Cuts
An Optimal Strategy for Computing Constrained Graph Cuts
A Practical Algorithm
Conclusions
References
Highly Consistent Sequential Segmentation
Introduction
Consistent Image Sequence Segmentation
Frame-to-Frame Segment Matching
Experiments
Evaluating Consistency
Object Tracking
Conclusion
References
Improved Video Segmentation by Adaptive Combination of Depth Keying and Mixture-of-Gaussians
Introduction
Segmentation
Segmentation by Depth Keying
Segmentation by Mixture-of-Gaussians
Combining Color and Depth
Results
Conclusions
References
Poster Session I
Sparse Similarity-Based Fisherfaces
Introduction
Algorithm Description
Obtaining the Texture Formulation
Creating the Individual Subspaces
Classification
Experimental Results
Classification Accuracy
Discriminative Pixels
Discussion and Conclusion
References
Accumulation of Different Visual Feature Descriptors in a Coherent Framework
Introduction
Feature Descriptors
The Accumulation Algorithm
Kalman Filtering
Accumulation of Confidence
Results
Conclusion
References
Person Re-identification by Descriptive and Discriminative Classification
Introduction
Person Re-identification System
Descriptive Person Model
Discriminative Person Model
Estimating Ranks by Boosting for Feature Selection
Features
Experimental Results
VIPeR Dataset
Multi-shot Dataset
Conclusion
References
On Inferring Image Label Information Using Rank Minimization for Supervised Concept Embedding
Introduction
Related Work
Method
Experiments
Data
Explicit Unsupervised Feature Extraction
Label Inference Results
Conclusion
References
Saliency in Spectral Images
Introduction
Saliency for Multispectral Images
Channels Computation
Center-Surround Comparisons
Experiments and Results
Data Sets
Saliency Map
Evaluation of Dimensionality Reduction Techniques
Conclusion
References
Mixed-State Particle Filtering for Simultaneous Tracking and Re-identification in Non-overlapping Camera Networks
Introduction
Learning Identities to Re-identify and Track
Target Representation
Reducing the Database to Key Frames
Embedding Re-identification into the Tracking Process
Particle Filter Framework
Estimating the Identity and the Position
The Non-ubiquity Constraint
Evaluations
Evaluation Network Setup
Re-Identification Efficiency
Multi-Target Re-identification
Conclusion and Perspectives
References
Smoothing-Based Submap Merging in Large Area SLAM
Introduction
Background and Related Work
SLAM Algorithms
Smoothing and Mapping
Submaps
Method
Conditionally Independent Submaps
Submap Merging
Examples
Implementation Details
Small Area Experiment
Large Area Experiment
Conclusions
References
Watermark Recovery from a Dual Layer Hologram with a Digital Camera
Introduction
Methods
Watermarking Method
Correcting Distortions and Segmentation
Experiments
Conclusion
References
Point Pattern Matching for 2-D Point Sets with Regular Structure
Introduction
Printed Electronics
Point Pattern Matching
Choosing Accumulator Array Size for Chang's Method
Modified Method without Discrete Accumulator Array
Example Case
Conclusions
References
Real Time Surface Registration for PET Motion Tracking
Introduction
Methods
Results and Discussion
Conclusion
References
Image Reconstruction by Prioritized Incremental Normalized Convolution
Introduction
Method
Local Polynomial Expansion: Assigning Values
Experiments and Results
Reconstruction of Randomly Removed Data
Reconstruction of Holes
Reconstruction of Missing 3D Surface Data
Discussion
Conclusion
References
Forming Different-Complexity Covariance-Model Subspaces through Piecewise-Constant Spectra for Hyperspectral Image Classification
Introduction
Model Formulation
Discriminant Analysis
Primary and Secondary Subspaces
Subspace Modeling Using PCA
Subspace Modeling Using Wider Spectral Bands
Primary-Subspace Size
Experiments
Data Sets
Experiment Details
Results
Discussion
Conclusion
References
Mobile Visual Search from Dynamic Image Databases
Introduction
Related Work
Mobile Image Matching from Dynamic Databases
Sub-linear Indexing
Descriptor Pruning
Matching with Multiple Indices
Experiments
Magazines Database
ZuBuD Database
Conclusions
References
Histogram-Based Description of Local Space-Time Appearance
Introduction
Related Work
Building the Spatio-temporal Appearance Descriptor
Spatio-temporal Appearance Descriptor of the First Order
Spatio-temporal Histogram Descriptor of the Second Order
Learning from the STA Descriptor
Entropy
The ?2 Measure
Using the STA Descriptor in Machine Learning
Illustrative Experiments
Discriminating between True and False Positives
Distinguishing between a Static and a Moving Background
Finding Stable Features to Track
Conclusion and Outlook
References
Content Based Detection of Popular Images in Large Image Databases
Introduction
The Image Database
Image Descriptors
Classification
Results
Summary and Conclusions
References
Unscented Kalman Filtering for Articulated Human Tracking
Introduction
Nonlinear Filtering
The Unscented Kalman Filter
The Particle Filter
Filtering for Articulated Tracking
The State Model
The Observation Model
Computational complexity
Results and Evaluation
Accuracy
Motion Smoothness
Conclusion
References
Using Fourier Descriptors and Spatial Models for Traffic Sign Recognition
Introduction
Related Work
Main Contribution
Methods
Fourier Descriptors
Matching of FDs
Sign Prototypes
Matching Sign Prototypes
Dataset
Experiments
Results Manually ROIs Dataset
Results Summer Dataset
Conclusions
References
Multichannel Segmentation Using Contour Relaxation: Fast Super-Pixels and Temporal Propagation
Motivation
The Region-Based Generative Image Model for Multi-channel Data
Segmentation as a Maximum-a-Posteriori (MAP) Estimation Problem
The Contour Relaxation
Likelihood Expressions for Region-Oriented Segmentation Approaches
The Data Likelihood Term for a Set of Measurements under the Gaussian Assumption
Experimental Results
Contour Relaxation Applied to Single Images
Temporal Propagation of the Segmentation
Conclusions
References
Color Persistent Anisotropic Diffusion of Images
Introduction
Related Work
Main Contribution
Methods
Standard Approach to Color Image Diffusion
Color Model
New Approach to Color Image Diffusion
Experiments
Data Set
Tested Methods
Performance Evaluation
Results
Conclusion
References
Analysis of Seed Sorting Process by Estimation of Seed Motion Trajectories
Introduction
Materials and Data
Experimental Setup
The Seed Material
Experiments and Acquired Data
Methods
Estimation of Seed Locations
Estimating the Angle of Escape
Results and Discussion
Conclusions
References
Improving Particle Segmentation from Process Images with Wiener Filtering
Introduction
Imaging and Image Processing
Wiener Filtering
Fiber Segmentation
Experiments and Discussions
Future Work
Conclusion
References
Efficient Hyperelastic Regularization for Registration
Introduction
Previous Work
Registration
Similarity Measure and Image Function
Regularization
Eigen Values and Derivatives
Choice of Basis Functions
Basis Examples
Experiments
Synthetic Data
Cardiac Data
Ear Data
Discussion
Summary and Conclusion
References
Image Analysis I
Degradation Based Blind Image Quality Evaluation
Introduction
Features for Detecting Degradations
Classifier for Quality Evaluation
Experimental Results
Features for Detecting Blur
Image Quality Classification
Conclusions
References
Evaluation of Image Quality Metrics for Color Prints
Introduction
Experimental Setup
Test Images
Printing Workflow
Observers
Viewing Conditions
Experiment Procedure
Experimental Results
Evaluation of Image Quality Metrics
Preparation of the Printed Images
Selected Image Quality Metrics
Evaluation Method
Evaluation Results
Conclusion and Future Work
References
Supercontinuum Light Sources for Hyperspectral Subsurface Laser Scattering Applications for Food Inspection
Introduction
Method
Experiments
Discussion
Conclusion
References
Categorization and Classification
Real-Time Detection of Landscape Scenes
Introduction
Methods for Landscape Scene Recognition
Global Features
Local Features
Classification
Experimental Results
Image Sets
Evaluation Criteria
Results
Computational Cost
Real-Time Implementation
Conclusion
References
Generic Object Class Detection Using Feature Maps
Introduction and Related Works
Object Class Model
Cascade
Strong Classifier
Weak Classifier
Single Feature Classifier
Variations
Detection
Hierarchical Search
Aspect Ratio Estimation
Experiments and Results
Conclusion
References
Volume Local Phase Quantization for Blur-Insensitive Dynamic Texture Classification
Introduction
Volume Local Phase Quantization
Short-Term Fourier Transform in the Spatio-Temporal Domain
Dimension Reduction
Quantization
Experiments
Classification Tests
Comparison on Execution Times
Conclusions
References
Structure from Motion and SLAM
Optimal View Path Planning for Visual SLAM
Introduction
Problem Formulation
Cost Function
Calculating Covariance
Cost Function
Cost Function Properties
Proposed Algorithm
Experiments
Discussion
Computational Complexity
Extensions
Conclusion
References
Automatic Estimation of the Number of Deformation Modes in Non-rigid SfM with Missing Data
Introduction
Deformable Low-Rank Shape Model
Proposed Approach
Frequency Content of 2D Trajectories
Algorithm
The reference Matrix
Experiments
Synthetic Data
Real Data
Conclusions
References
Unsupervised Learning for Improving Efficiency of Dense Three-Dimensional Scene Recovery in Corridor Mapping
Introduction
Dense and Efficient 3D Scene Recovery
Recovering Planar Structures
Unsupervised Learning for Temporal Search Space Reduction
Experimental Results
Conclusions
References
Catadioptric Silhouette-Based Pose Estimation from Learned Models
Introduction
Method
Silhouette Extraction
Catadioptric Camera Setup and Calibration
Reconstruction Refinement
Model-Based Pose Estimation
Experiments
Calibration Accuracy
Pose Estimation with Synthetic Images
Pose Estimation with Real Images
Conclusion
References
Medical and Biomedical Applications
Using the Local Phase of the Magnitude of the Local Structure Tensor for Image Registration
Introduction
Methods
Quadrature Filters and Local Phase
The Local Structure Tensor
Local Phase of the Magnitude of the Local Structure Tensor
Results
Discussion
References
Coherence Probe Microscopy Imaging and Analysis for Fiber-Reinforced Polymers
Introduction
Methods
CPM Measurement Setup and Materials
Image Demodulation
Image Enhancement
Orientation Estimation
Results
Image Enhancement
Orientation Estimation
Clustering
Conclusions
References
Automatic Segmentation of Veterinary Infrared Images with the Active Shape Approach
Motivation and State of the Art
Image Acquisition and IR Temperature Calculation
IRT Segmentation and the Active Shape Approach
Choice of Segmentation Criterion
Specific Modifications of the ASM Segmentation
Image Preprocessing
Landmark Representation
Shape Creation and Initialisation
Application Results
Conclusions and Future Work
References
3D Shape
Using the Skeleton for 3D Object Decomposition
Introduction
Preliminaries
Object Decomposition
Skeleton Partition
Recovery of Simple Regions, Bumps and Kernels
Improving the Separation between Parts
Removal of Sub-parts
Expanding Simple Regions, Bumps and Kernels
Merging
Discussion and Conclusion
References
A Three-Dimensional Shape Description Algorithm Based on Polar-Fourier Transform for 3D Model Retrieval
Introduction
Description of the Polar-Fourier 3D Shape Descriptor
Brief Description of the Algorithms Selected for the Comparison with the Proposed Method
Experimental Conditions and Results
Conclusions and Future Plans
References
Combining Stereo and Time-of-Flight Images with Application to Automatic Plant Phenotyping
Introduction
Methods
Setup and Calibration
Dense Stereo Methods
Localised Search Range from ToF Image
Quality Quantification
Results
Discussion
References
Medical Imaging
Iterative Reconstruction for Quantitative Tissue Decomposition in Dual-Energy CT
Introduction
Information Attainable from CT Imaging
Recent Technical Developments in CT Imaging
Tissue Classification Using Single-Energy CT Scans
Tissue Classification Using Dual-Energy CT Scans
Methods
Filtered Backprojection Reconstruction in CT
Segmentation of Body Tissues
The Three Material Decomposition method
The Iterative Reconstruction Algorithm
Experiments
Considerations for the Iterative Reconstruction Algorithm
The Mathematical Phantom
Linear Attenuation Coefficients and Effective Energies
Results
Conclusions and Future Work
References
An Automated System for the Detection and Diagnosis of Kidney Lesions in Children from Scintigraphy Images
Introduction
Clinical Background
Related Work
Methods
Kidney Segmentation
Boundary Representation
Background Removal
Candidate Lesion Segmentation
Classification
Results
Discussion
References
Automatic Segmentation of Abdominal Adipose Tissue in MRI
Introduction
Data
Methodology
Bias Field Correction
Identifying Adipose Tissue
Identifying Regions
Results
Conclusion
References
Fully Automatic Liver Volumetry Using 3D Level Set Segmentation for Differentiated Liver Tissue Types in Multiple Contrast MR Datasets
Introduction
Related Work
Data Acquisition
Description of the Method
Denoising of MR Datasets
Training Phase
Probability Map Generation
3D Level Set Segmentation
Distance Transform Based Segmentation Refinement
Results
Discussion
References
Poster Session II
Stable Structure from Motion for Unordered Image Collections
Introduction
System Overview
Point Tracking in Unordered Image Collections
Robust Global Rotation Computation
Global Geometry Estimation
Results and Conclusions
References
Projector Calibration by "Inverse Camera Calibration"
Introduction
Related Work
Projector Calibration
Camera Calibration
Inverting Camera Calibration
Proposed Calibration Method
Experiments
Conclusion
References
Representing Local Structure Using Tensors II
Introduction
Monomial Filters
Signal Classes
Second Order Structure Tensors
Structure Tensor Estimation Variations
Higher Order Structure Tensors
Conclusion
References
Automatic Compartment Modelling and Segmentation for Dynamical Renal Scintigraphies
Introduction
Methods
Using SVD to Approximate the Data
Using Classification to Approximate the Data
Estimation of bk
Estimation of ck Using Positivity Constraints and Spatial Smoothness Priors
Compartment Modelling
Incorporating All Constraint Modalities
Experimental Validation
Discussion and Conclusions
References
Expression Recognition in Videos Using a Weighted Component-Based Feature Descriptor
Introduction
Component-Based Feature Descriptor
Dynamic Texture Features For Appearance Representation
Dynamic Structural Features For Shape Representation
Multiple Feature Fusion
The Weight Learning Framework
The Fusion Module Framework
Experiments
Conclusion
References
Spatio-chromatic Image Content Descriptors and Their Analysis Using Extreme Value Theory
Introduction
Spatio-chromatic Descriptors
Extreme Value Theory
The Basics of Univariate EVT
A Simple Stochastic Model
Proposed Approach
Experiments
Statistical Analysis: Goodness of Fit
Further Analysis: The s, k-Space
Classification and Retrieval
Conclusion
References
Model-Based Transfer Functions for Efficient Visualization of Medical Image Volumes
Introduction
Model-Based Local Transfer Function
Model
Registration
Visualization Based Upon Local Transfer Functions
Smooth Transition Regions between Transfer Functions
Evaluation
Implementation
Results
Results from the Evaluation
Results from the Implementation
Discussion
References
Anatomical Landmark Tracking for the Analysis of Animal Locomotion in X-ray Videos Using Active Appearance Models
Introduction
Related Work and Motivation
Active Appearance Models
Training Step
Model Fitting
Application to X-ray Locomotion Landmark Tracking
Experiments and Results
General Suitability
Impact of Image Resolution and Preprocessing
Selection of Training Images
Conclusions and Further Work
References
Image Foresting Transform: On-the-Fly Computation of Segmentation Boundaries
Introduction
Background
Images and Graphs
Paths and Path Costs
Spanning Forests
Image Segmentation
The Image Foresting Transform
The Differential Image Foresting Transform
Applications
Sub-pixel Segmentation with the IFT
The Relaxed IFT
Conclusion
References
Scale Space Smoothing, Image Feature Extraction and Bessel Filters
Introduction
Theory
Implementation Issues
Numerical Results
Conclusion
References
A Free-Viewpoint Virtual Mirror with Marker-Less User Interaction
Introduction
Related Work
System Design
Geometric Considerations for the Mirror
System Architecture
Methodology
3D Reconstruction and Robust Skeleton Estimation
Hand Position Estimation
Virtual Mirror Camera
User Interaction
Image-Based Visual Hull Rendering
Experiments
User Interaction
Run-Time Evaluation
Conclusions and Discussions
References
Wood Detection and Tracking in Videos of Rivers
Introduction
Related Works
Limitation of Existing Methods in Our Case
Segmentation of Floating Wood
Intensity Probability Map
Temporal Probability Map
Combination of Intensity and Temporal Probability Maps
Selection of Parameters
Tracking of Floating Wood
Extraction of Representative Points
Temporal Linking of Floating Wood
Counting Wood Pieces
Experimental Results
Concluding Remarks
References
Interactive Image Segmentation Using Level Sets and Dempster-Shafer Theory of Evidence
Introduction
Level Set Segmentation with Evidence Theory
Interactive Variational Image Segmentation
Experiments
Conclusion
References
Fast and Efficient Saliency Detection Using Sparse Sampling and Kernel Density Estimation
Introduction
Saliency Measurement
Bayesian Center-Surround
Defining Saliency Measure
Multi-scale Measure
Implementation
Experiments
Quantitative Analysis
Qualitative Assessment
Conclusion
References
Combining Contrast Information and Local Binary Patterns for Gender Classification
Introduction
Face Representation Using Local Binary Patterns
Face Representation Using LBP and Contrast
Experiments on Gender Classification
Experimental Data
Settings
Results and Analysis
Conclusion
References
Indexing Tree Structures through Caterpillar Decomposition
Introduction
Preliminaries
Indexing Tree Structures through Caterpillar Decomposition
Experiments
Conclusions
References
Recovering Missing Data on Satellite Images
Introduction
Variational Data Assimilation
Mathematical Setting
Variational Formulation
Recovering of Missing Data
Results
Artificial Noise
Real Noise
Conclusion
References
Target Segmentation in Scenes with Diverse Background
Introduction
Clustering of Object and Background Data
Gaussian Mixture Model
The Minimum Message Length Criterion
The EM-MML Algorithm and Initialization
The Normalized Scatter Separability Criterion
The Spatial Scatter Criterion
Feature Sorting
Examples
Sensor Data and Data Registration
Features
Parameter Settings
Results
Conclusions and Future Work
References
A Hybrid Approach to Brain Extraction from Premature Infant MRI
Introduction
Proposed Method
Overview
Data Pre-processing
Creating Mask
Automatic Seedpoint Detection
ROI Segmentation
Evaluation
Data Aquisition
Visual Examination
Similarity Metrics and Numerical Evaluation
Discussion
Conclusion
References
Adaptive Classification of Dirt Particles in Papermaking Process
Introduction
Classification of Dirt Particles
Problem Statement
General Workflow
Grouping of Dirt Particles Based on a Size
Segmentation
Feature Extraction
Feature Evaluation
Classification
Experiments and Discussion
Test Samples and Experiments
Classification on the Whole Set
Classification Using Subsets
Conclusion
References
Text Extraction Using Component Analysis and Neuro-fuzzy Classification on Complex Backgrounds
Introduction
Description of the Technique
Color Reduction
Adaptive Run Length Smoothing Algorithm
Image Pre-processing
Document Segmentation
Feature Extraction
Pattern Classification Using ANFIS
Experimental Results
Conclusions
References
Using Active Illumination for Accurate Variational Space-Time Stereo
Introduction
The Space-Time Variational Method
Active Illumination
Evaluation
Conclusion
References
Kernel Fisher Discriminant and Elliptic Shape Model for Automatic Measurement of Allergic Reactions
Introduction
Grayscale Projection
Elliptic Shape Model
Experimental Results
Conclusions
References
Image Analysis II
Continuous Orientation Representation for Arbitrary Dimensions - A Generalized Knutsson Mapping
Introduction
Construction of the General Knutsson Mapping
The 2D Case in the General Framework
The 3D Case in the General Framework
The $n$D Case in the General Framework
Alternative Mapping Based on Projection Operators and Angles between Vectors in Mapped Space
Preserving Isotropy in Mapped Space
Doubling the Angle - Knutsson Cones in Higher Dimensions
Conclusion
References
Real-Time Line Detection Using Accelerated High-Resolution Hough Transform
Introduction
Real-Time Hough Transform Algorithm
Hough Transform on a Small Read-Write Memory of Accumulators
Harnessing the Edge Orientation
Experimental Results
Performance Evaluation on Synthetic Binary Images
Performance Evaluation on Real-Life Images
Conclusions
References
Decomposition of a Curve into Arcs and Line Segments Based on Dominant Point Detection
Introduction
Dominant Point Detection
Blurred Segment
A Method for Dominant Point Detection
Arc Detection
Tangent Space Representation
Properties of Arc in the Tangent Space
Algorithm for Arc (Circle) Detection
Curve Decomposition into Arcs and Lines
Main Idea of the Proposed Method
Analysis of Configurations
Proposed Algorithm
Experimentations
Experimental Results and Comparisons
Application to Vectorization
Conclusion
References
Author Index

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Image Analysis

Description

More details

Other editions

Additional editions

Content

System requirements