Advanced Concepts for Intelligent Vision Systems

Name: Advanced Concepts for Intelligent Vision Systems | 13th International Conference, ACIVS 2011, Ghent, Belgium, August 22-25, 2011, Proceedings
Brand: Springer
Price: 53.49 EUR
Availability: OnlineOnly

13th International Conference, ACIVS 2011, Ghent, Belgium, August 22-25, 2011, Proceedings

Jaques Blanc-Talon Wilfried Philips Dan Popescu Paul Scheunders Richard Kleihorst(Editor)

Springer (Publisher)

Published on 6. September 2011

XVIII, 763 pages

E-Book

PDF with digital watermarking

System requirements

978-3-642-23687-7 (ISBN)

€53.49incl. 7% vat

System requirements

for PDF with digital watermarking

E-Book Single Licence

Available for download

Description

More details

Other editions

Content

Title
Preface
Organization
Table of Contents
Vision
Robust Visual Odometry Using Uncertainty Models
Introduction
Camera Model
The Proposed Method
Results
Conclusion
References
Supervised Visual Vocabulary with Category Information
Introduction
Supervised Vocabulary Construction
Supervised Vocabulary Construction
Experiments
Experimental Setup
Experimental Results
Conclusions
References
Nonparametric Estimation of Fisher Vectors to Aggregate Image Descriptors
Introduction
Aggregating the Image Feature
Fisher Kernel, Score and Vector
Logspline Density Estimation
Signature Derivation
Signature Normalization
Efficient Implementation
Independent Low Level Features
Experiments
Datasets and Experimental Setup
Signature Implementation
Signature Parameters
Comparison to Other Works
Conclusion
References
Knowledge-Driven Saliency: Attention to the Unseen
Introduction
Related Literature
Proposed System
Multisensor Input and Preprocessing
View Descriptors
3D Object Construction Using Views
Learning of Objects into Knowledge
Object Recognition in 3D
Learning of 3D Scenes
Knowledge-Driven Saliency
Results
Conclusion
References
A Comparative Study of Vision-Based Lane Detection Methods
Introduction
State of the Art
Model-Based Approach
Feature-Based Approach
A New Method for Lane Detection
Pre-processing
Extraction of Approximate Pixels
Lane Detection
Experimental Results and Performance Evaluation
The Test Dataset
Experimental Results
Performance Evaluation
Conclusion
References
A New Multi-camera Approach for Lane Departure Warning
Introduction
Pre-processing
Camera Calibration
Height Measurement
Offset between Cameras
Distance between Cameras
Lane Detection
Data Fusion
Lane Departure Warning
Results and Analysis
Conclusion
Future Work
References
Classification, Recognition and Tracking
Feature Space Warping Relevance Feedback with Transductive Learning
Introduction
Related Work
Visual Similarity Using Covariance Matrices
Mean-Shift Feature Space Warping with Remapping
Transductive Relevance Feedback
Transductive Relevance Feedback with Feature Space Warping
Experimental Results
Conclusions
References
Improved Support Vector Machines with Distance Metric Learning
Introduction
Support Vector Machine with Metric
Distance Metric Learning
Experiments
Low Dimension Data Sets
Image Data Sets
Conclusions
References
A Low-Cost System to Detect Bunches of Grapes in Natural Environment from Color Images
Introduction
The Grape Recognition System
Results and Discussion
Conclusions and Ongoing Work
References
Fuzzy Cognitive Maps Applied to Synthetic Aperture Radar Image Classifications
Introduction
Fuzzy Cognitive Maps Framework in SAR Classification
The Combined H -a Decomposition and Maximum Likelihood-Based Wishart
Cluster Separation Measures
The Fuzzy Cognitive Maps Process
Results
Design of a Test Strategy
Results
Discussion
Conclusions
References
Swarm Intelligence Based Searching Schemes for Articulated 3D Body Motion Tracking
Introduction
Searching Schemes for Articulated Human Body Tracking
Tracking Framework
Experimental Results
Conclusions
References
Combining Linear Dimensionality Reduction and Locality Preserving Projections with Feature Selection for Recognition Tasks
Introduction
Linear Dimensionality Reduction
Principal Component Analysis (PCA)
Locality Preserving Projections (LPP)
PCA+LPP
Feature Selection
Feature Selection
Eigenvector Selection for Object Recognition
Experimental Results
Benchmark Data Sets
Method Comparison
Conclusion
References
A New Anticorrelation-Based Spectral Clustering Formulation
Introduction
Development of Our Method
Results of Comparison to Kernel-PCA
Connection of SCE with Normalized Cuts
SCE Extension with Latent Variables
Conclusions and Future Work
References
Simultaneous Partitioned Sampling for Articulated Object Tracking
Introduction
Reducing Particle Filter's Complexity: The Problem of High-Dimentional State Spaces
Proposed Approach
Partitioned Sampling (PS)
Our Contribution: Simultaneous Partitioned Sampling (SPS)
Experimental Results
Conclusion
References
Segmentation
A Geographical Approach to Self-Organizing Maps Algorithm Applied to Image Segmentation
Introduction
Literature Review
SOM Applied to Remote Sensing
Segmentation Using SOM
Methodology
Results
Conclusion
References
A Multi-Layer 'Gas of Circles' Markov Random Field Model for the Extraction of Overlapping Near-Circular Objects
Introduction
The Single-Layer `Gas of Circles' Model
Contour Representation
Phase Field Representation
Binary MRF Representation
The Multi-Layer MRF `Gas of Circles' Model
Energy of Two Interacting Circles
Experimental Results
Data Likelihood
Simulation Results with the Multi-Layer MRF GOC Model
Quantitative Evaluation on Synthetic Images
Application in Biomedical Imaging
Conclusion
References
Evaluation of Image Segmentation Algorithms from the Perspective of Salient Region Detection
Introduction
Related Work
Segmentation Methods
Graph-Based Salient Object Detection
Unsupervised Segmentation of Natural Images via Lossy Data Compression
ROI-SEG: Unsupervised Color Segmentation by Combining Differently Focused Sub Results
Multi-Layer Spectral Segmentation
Efficient Graph-Based Image Segmentation
Normalized Cuts
Mean Shift
Saliency Detection Performance Evaluation
Experimental Results
Conclusion and Future Work
References
Robust Active Contour Segmentation with an Efficient Global Optimizer
Introduction
Notations and Definitions
Convex Energy Active Contours
Optimization
Results
Examples
Error Metric
Noise Robustness
Conclusion
References
A Method to Generate Artificial 2D Shape Contour Based in Fourier Transform and Genetic Algorithms
Introduction
A Brief Theoretical Introduction
Shape Representation in the Frequency Domain
Genetic Algorithms
The Proposed Method
Validation Method
Results and Conclusions
References
Image Segmentation Based on Electrical Proximity in a Resistor- Capacitor Network
Introduction
Proximity in a Resistor-Capacitor Network
Using the Proximity for Image Segmentation
Experimental Results
Conclusion
References
Hierarchical Blurring Mean-Shift
Introduction
Mean-Shift
Blurring Mean-Shift
Hierarchical Blurring Mean-Shift
Experiments
Conclusion
References
Image Analysis
Curve-Skeletons Based on the Fat Graph Approximation
Introduction
Definitions
Fat Graph Fitting
General Overview
Fat Graph in the Fixed Topological Class
Implementation
Experiments
Conclusion
References
DTW for Matching Radon Features: A Pattern Recognition and Retrieval Method
Introduction
Method
Pattern Representation
Pattern Matching
Pattern Recognition and Retrieval
Experiments
Benchmarking Methods
Experimental Results
Discussions
Conclusions
References
Ridges and Valleys Detection in Images Using Difference of Rotating Half Smoothing Filters
Introduction
A Rotating Smoothing Filter
Ridge/Valley Lines Detection Using Difference of Directional Gaussian Filters
Difference of Rotated Half Smoothing Filters (DRF)
Pixel Classification
Ridge and Valley Extractions
Results
Results on a Synthetic Image
Results on Real Images
Results Evaluation
Conclusion
References
Analysis of Wear Debris through Classification
Introduction
Background
Description of Data
Normalization
Classification
Image Features
Machine Learning
Experiments and Results
Experimental Setup
Results
Conclusions
References
Fourier Fractal Descriptors for Colored Texture Analysis
Introduction
Fractal Geometry
Fractal Dimension
Fourier Fractal Dimension
Fractal Descriptors
Proposed Method
Experiments
Results
Conclusion
References
Efficiency Optimization of Trainable Feature Extractors for a Consumer Platform
Introduction
CNN Algorithm Overview
Algorithm Optimization
Merge Convolution and Subsampling
Training with Error Back-Propagation
Validate the Recognition Performance
Implementation
Related Work
Conclusion
References
Salient Region Detection Using Discriminative Feature Selection
Introduction
Salient Region Detection
Identification of Photographic Composition
Feature Selection for Saliency
Computing Saliency
Experimental Results
Conclusions
References
Image Analysis Applied to Morphological Assessment in Bovine Livestock
Introduction
First Stage: Image Analysis
Shape Modelling and Search Space
Objective Function and Potential Image Extraction
Genetic Algorithm Configuration
Fixing Points to Drive the Search: Method
Second Stage: Neural Network
Experiments and Results
First Stage
Second Stage: Assessment
Conclusions and Future Research
References
Quantifying Appearance Retention in Carpets Using Geometrical Local Binary Patterns
Introduction
Materials
Methods
Experiment
Results
Conclusions
References
Enhancing the Texture Attribute with Partial Differential Equations: A Case of Study with Gabor Filters
Introduction
Background
Anisotropic Diffusion
Gabor Filters
An Approach to Texture Analysis
Experimental Results
Datasets
Performance Evaluation
Conclusions
References
Dynamic Texture Analysis and Classification Using Deterministic Partially Self-avoiding Walks
Introduction
Deterministic Partially Self-avoiding Walks
Modeling Dynamic Textures
Deterministic Partially Self-avoiding Walks on Sequence of Images
Feature Vector
Results and Discussion
Comparison with other Methods
Clustering Traffic Database
Conclusion
References
Image Processing
Segmentation Based Tone-Mapping for High Dynamic Range Images
Introduction
Review of Tone Mapping Methods
Algorithm
Image Segmentation
Local Tone Mapping in Segmented Regions
Eliminate the Boundary Artifacts
Adaptive Selection of Parameter ß in Uniform Areas
Experimental Results and Discussion
Mapping Results and Discussion
Computational Efficiency
Conclusion and Future Work
References
Underwater Image Enhancement: Using Wavelength Compensation and Image Dehazing (WCID)
Introduction
Underwater Image Model
Distance between the Camera and the Object: d(x)
Underwater Depth of the Photographic Scene D
Image Depth Range R
Artificial Light Source L
Conclusion
References
Video Stippling
Introduction
Voronoi Diagrams
CVT Method for Video Stippling
From Images to Video
Image Differences and Drift Adjustment of Sites
Color and Size
Frequency Considerations
Pattern placement.
Experiments and Results
Blue Noise Characteristics
Algorithm's Complexity
Size of the Output File
Discussion and Concluding Remarks
References
Contrast Enhanced Ultrasound Images Restoration
Introduction
Noise Removal
Our Approach
Results
Conclusion
References
Mutual Information Refinement for Flash-no-Flash Image Alignment
Introduction
Alignment Methods
Feature Based Methods
Image Based Methods
Proposed Method
Results and Discussion
Effects of the Window Sizes
Comparison to State-of-the-Art Methods
Conclusions and Discussion
References
Virtual Restoration of the Ghent Altarpiece Using Crack Detection and Inpainting
Introduction
Crack Detection
Multiscale Crack Map Formation
Crack Map Refinement
Crack Inpainting
Patch-Based Inpainting
Improved Patch-Based Inpainting for Crack Filling
Conclusion
References
Image Sharpening by DWT-Based Hysteresis
Introduction
The Proposed Method
Experimental Results
Conclusion
References
Content Makes the Difference in Compression Standard Quality Assessment
Introduction
Experimental Protocols
Experimental Results
Conclusion
References
A Bio-Inspired Image Coder with Temporal Scalability
Introduction
Virtual Retina: A Bio-Inspired Retina Model
The Coding Pathway
The Image Transform: The Outer Retina Layers
The A/D Converter: Inner and Ganglionic Layers
The Decoding Pathway
Results
Conclusion
References
Self-Similarity Measure for Assessment of Image Visual Quality
Introduction
Self-Similarity Measure
Basic Definitions
Calculation of Correcting Factors P and T
Accounting for Human Visual System
Visual Quality Metric Based on the Proposed Self-Similarity Parameter
Analysis of MSDDM Efficiency for Specialized Databases
Analysis of MSDDM Efficiency for Different Types of Distortions in TID
Conclusions
References
Video Surveillance and Biometrics
An Intelligent Video Security System Using Object Tracking and Shape Recognition
Introduction
Overall Structure
Object Extraction
Background Subtraction
Boundary Smoothing
Human Recognition
Overview of ART
Construction of Database
Modeling of ART Vectors
Face Detection
Human Tracking
Experimental Results
Conclusion
References
3D Facial Expression Recognition Based on Histograms of Surface Differential Quantities
Introduction
Estimating Differential Quantities on Triangular Meshes
Estimating Curvature by Normal Cycle Theory Based Method
Estimating Curvature by Local Cubic Fitting Based Method
Mesh Gradient Estimation
Local Shape Descriptors
Landmarks Selection
Local Coordinate System and Orientation Assignment
Feature Vectors Computed as Histograms of Surface Differential Quantities
Experimental Results
Conclusion and Future Work
References
Facial Feature Tracking for Emotional Dynamic Analysis
Introduction
Facial Features Tracking Method
Overview
Features
Multi-Kernel Learning
Statistical Validation
Experimental Setup
Training Database
Multi-Kernel Learning
Testing Phase
Experiments on the Cohn-Kanade Database
Performance Measures
Experimental Results
Sensitivity to Initialization
Temporal Analysis of Expressions
MMI Database
Temporal Segment Detection
Segmentation of Basic Emotions
Actions Units Segmentation: Comparison with Other Works
Conclusion
References
Detection of Human Groups in Videos
Introduction
Framework Overview and Group Detection Algorithm
Pre-processing
Multi-target Tracking
Human Group Detection Algorithm
Experiments and Results
Conclusions and Future Work
References
Estimation of Human Orientation in Images Captured with a Range Camera
Introduction
Related Works
Our Definition of the Orientation
Criterion Used for Evaluation
The Intrinsic Limitation of a Color Camera
Our Method
Data
Silhouette Description
Regression Method
Experiments
The Impact of Perspective Effects
The Role Played by 3D in Orientation Estimation
Observations for a Practical Application in Real Time
Conclusions
References
Human Identification Based on Gait Paths
Introduction
Motion Capture
Related Work
Collected Database of Human Gaits
Experiments, Results and Conclusions
References
Separating Occluded Humans by Bayesian Pixel Classifier with Re-weighted Posterior Probability
Introduction
Separating Occluded Humans by Bayesian Pixel Classifier
Training the Bayesian Pixel Classifier
Re-weight Posterior Probability
Separating Occluded Humans Using a Bayesian Pixel Classifier
Experimental Results
Conclusion and Future Work
References
An Edge-Based Approach for Robust Foreground Detection
Introduction
Background Subtraction Using Moving Edges
Short-Term Model
Long-Term Background Edge Model
Detection of Moving Edges
Results and Discussion
Visual Evaluation of the Proposed Method under Different Lighting Conditions
Numerical Evaluation of the Proposed Method for the Construction of Occupancy Maps
Conclusion
References
Relation Learning - A New Approach to Face Recognition
Introduction
Distance-Based Classification Methods
Proposed Relation Learning Method
Relation Learning for Face Recognition
FERET Evaluation Protocol
Relation Function for Training Set
Experiments on Face Recognition
Conclusion
References
Algorithms and Optimizations
Temporal Prediction and Spatial Regularization in Differential Optical Flow
Introduction
Bayesian Multi-scale Differential Optical Flow
Temporal Prior
Generating the Prediction
Incorporating Predictions
Results
Conclusion
References
Parallel Implementation of the Integral Histogram
Introduction
Computation of the Integral Histogram
Parallelization Using Tiles and Scan Propagation
Star Superscalar (StarSs) Parallel Programming Model
Implementation Using Tile-Level Halos
Experiments
Integral Histogram for Single Images
Integral Histogram for Sequences of Images
Conclusions and Future Work
References
System on Chip Coprocessors for High Speed Image Feature Detection and Matching
Introduction
Description of the Implemented Algorithms
FAST Feature Detector
SURF Feature Descriptor
Description of the Implemented Coprocessors
Feature Detection and Description Processor
Feature Matching Processor
Performance Evaluation
Conclusions
References
Fast Hough Transform on GPUs: Exploration of Algorithm Trade-Offs
Introduction
Hough Transform for Lines
Cartesian Hough Transform
Polar Hough Transform
Benchmark Setup
GPU Implementations
GPU Implementation 1 - Basic
GPU Implementation 2 - Fast
GPU Implementation 3 - Constant
Results
Related Work
Conclusion
References
Feasibility Analysis of Ultra High Frame Rate Visual Servoing on FPGA and SIMD Processor
Introduction
Visual Servoing System and Algorithm Development
Feasibility Analysis of FPGA Realization
Feasibility Analysis of SIMD Realization
Proposed Reconfigurable Wide SIMD Architecture
Visual Servoing on the Proposed Wide SIMD Processor
Conclusions and Future Work
References
3D, Depth and Scene Understanding
Calibration and Reconstruction Algorithms for a Handheld 3D Laser Scanner
Introduction
Sensor Geometry
Uncertainty in Sensor Geometry
Calibration and Reconstruction
Experimental Results
Conclusions and Future Work
References
Comparison of Visual Registration Approaches of 3D Models for Orthodontics
Introduction
Model Registration from Mouth Images
Estimation of Pi With K Knowledge (WiK)
Etimation of Pi Without K Knowledge (WoK)
Registration of the Mandible
Experiments
Projection Matrix Estimation
Mandible Registration
Conclusion
References
A Space-Time Depth Super-Resolution Scheme for 3D Face Scanning
Introduction
Offline and Online Preprocessing
3D Sparse Model and Projector Axis Estimation
3D Spatial Super-Resolution
3D Space-Time Super-Resolution
Experimental Results
Conclusion and Future Work
References
Real-Time Depth Estimation with Wide Detectable Range Using Horizontal Planes of Sharp Focus Proceedings
Introduction
The Drawback of the Conventional Method with a Single Camera for Real-Time Use
Depth Estimation by the DOF of a Tilted Optical Lens
A Tilted POF
DOF by a Tilted POF
Depth Estimation by a Horizontal POF
Depth Estimation by Using DOF
Our Proposal for Depth Estimation Using Horizontal POFs
Sharpness Curve with Horizontal POFs
Three Horizontal POFs and Sharpness Curves
Algorithm of Our Proposed Method
Our Proposed Method and the DFD Method
Experiment with Our Proposed Method
Experiment Environment
Relationship between Depth Value and Sharpness Curve
The Experiment for Our Proposed Method
Conclusion
References
Automatic Occlusion Removal from Facades for 3D Urban Reconstruction
Introduction
Previous Work
Cognitive 3D
Occlusion Removal
Geometric Analysis
3D Reconstruction with ARC3D
Plane Extraction
Image Rectification
Semantic Analysis
Vehicle Detection
Vehicle Segmentation
Vegetation Detection
Occlusion Removal
Inpainting
Removing Occlusions in 3D
Mesh Generation
Results
Planar Constraint on Inpainting
Mask Selection
Conclusion
References
hSGM: Hierarchical Pyramid Based Stereo Matching Algorithm
Introduction
Related Work
Semi-Global Optimization on the Hierarchical Pyramid
Propagation of Aggregated Matching Costs
The GPU Implementation
Experimental Result
Conclusion and Future Work
References
Surface Reconstruction of Rotating Objects from Monocular Video
Introduction
Previous Work
Method
Experimental Results
Conclusions
References
Precise Registration of 3D Images Acquired from a Hand-Held Visual Sensor
Introduction
The Sensor
Registration
Image-Based Registration
Image-Based Pairing
Transformation Estimation
Pyramidal Approach
Results
Mechanical Object
Cylinder Sequence
Conclusion
References
A 3-D Tube Scanning Technique Based on Axis and Center Alignment of Multi-laser Triangulation
Introduction
Overview of a Tube Scanning Module
Camera-Laser Scanning Module
Camera and Laser Calibration
A Circular Shape Reconstruction by Multi-laser Triangulation
Laser Curve Extraction and Fitting
3D Reconstruction of Laser Curves
Complete Tube Scanning Based on Axis and Center Alignment
Calculation of Tube Center and Axis
Tube Center and Axis Alignment
Complete Tube Reconstruction
Experiment Results
Experiment 1
Experiment 2
Conclusion and Future Works
References
Combining Plane Estimation with Shape Detection for Holistic Scene Understanding
Introduction
Background and Related Work
Basic Shape and Plane Estimation
Basic Shape Estimation Based on Perceptual Organization
Plane Estimation
Joint Probabilistic Model
Experiments
Conclusion
References
Simple Single View Scene Calibration
Introduction
Related Work
Proposed Calibration Method
Projection Model
Finding Optimal Solution
Experimental Results
Conclusions
References
Author Index

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Advanced Concepts for Intelligent Vision Systems

Description

More details

Other editions

Additional editions

Content

System requirements