Computer Vision for X-Ray Testing Imaging, Systems, Image Databases, and Algorithms

This accessible textbook presents an introduction to computer vision algorithms for industrially-relevant applications of X-ray testing. Covering complex topics in an easy-to-understand way, without requiring any prior knowledge in the field, the book provides a concise review of the key methodologi...

Full description

Saved in:
Bibliographic Details
Main Authors Mery, Domingo, Pieringer, Christian
Format eBook
LanguageEnglish
Published Cham Springer International Publishing AG 2020
Springer International Publishing
Edition2
Subjects
Computer Science
Computer vision
Image Processing and Computer Vision
Machine Learning
Quality Control, Reliability, Safety and Risk
Simulation and Modeling
Online AccessGet full text
ISBN9783030567682
3030567680
DOI10.1007/978-3-030-56769-9

Cover

Table of Contents:
  • Intro -- Foreword to the Second Edition -- Foreword to the First Edition -- Preface to the Second Edition -- Preface to the First Edition -- Acknowledgements -- Contents -- About the Authors -- 1 X-ray Testing -- 1.1 Introduction -- 1.2 History -- 1.3 Physics of the X-rays -- 1.3.1 Formation of X-rays -- 1.3.2 Scattering and Absorption of X-rays -- 1.4 X-ray Testing System -- 1.4.1 X-ray Source -- 1.4.2 Manipulator -- 1.4.3 Image Intensifier -- 1.4.4 CCD-Camera -- 1.4.5 Flat Panel -- 1.4.6 Computer -- 1.5 X-ray Imaging -- 1.5.1 X-ray Image Formation -- 1.5.2 Image Acquisition -- 1.5.3 X-ray Image Visualization -- 1.5.4 Dual-Energy -- 1.6 Computer Vision -- 1.6.1 Geometric Model -- 1.6.2 Single View Analysis -- 1.6.3 Multiple View Analysis -- 1.6.4 Deep Learning -- 1.6.5 Computed Tomography -- 1.7 Code and Data -- 1.7.1 Pyxvis Library -- 1.7.2 mathbbGDXray+ Database -- 1.8 General Methodology for X-ray Testing -- 1.9 Summary -- References -- 2 Images for X-ray Testing -- 2.1 Introduction -- 2.2 Structure of the Database -- 2.3 Castings -- 2.4 Welds -- 2.5 Baggage -- 2.6 Natural Objects -- 2.7 Settings -- 2.8 Python Commands -- 2.9 Summary -- References -- 3 Geometry in X-ray Testing -- 3.1 Introduction -- 3.2 Geometric Transformations -- 3.2.1 Homogeneous Coordinates -- 3.2.2 2D rightarrow 2D Transformation -- 3.2.3 3D rightarrow 3D Transformation -- 3.2.4 3D rightarrow 2D Transformation -- 3.3 Geometric Model of an X-ray Computer Vision System -- 3.3.1 A General Model -- 3.3.2 Geometric Models of the Computer Vision System -- 3.3.3 Explicit Geometric Model Using an Image Intensifier -- 3.3.4 Multiple View Model -- 3.4 Calibration -- 3.4.1 Calibration Using Python -- 3.4.2 Experiments of Calibration -- 3.5 Geometric Correspondence in Multiple Views -- 3.5.1 Correspondence Between Two Views -- 3.5.2 Correspondence Between Three Views
  • References -- 6 Classification in X-Ray Testing -- 6.1 Introduction -- 6.2 Classifiers -- 6.2.1 Minimal Distance -- 6.2.2 Mahalanobis Distance -- 6.2.3 Bayes -- 6.2.4 Linear Discriminant Analysis -- 6.2.5 Quadratic Discriminant Analysis -- 6.2.6 K-Nearest Neighbors -- 6.2.7 Neural Networks -- 6.2.8 Support Vector Machines -- 6.2.9 Classification Using Sparse Representations -- 6.3 Performance Evaluation -- 6.3.1 Hold-Out -- 6.3.2 Cross-Validation -- 6.3.3 Leave-One-Out -- 6.3.4 Confusion Matrix -- 6.3.5 ROC and Precision-Recall Curves -- 6.4 Classifier Selection -- 6.5 Summary -- References -- 7 Deep Learning in X-ray Testing -- 7.1 Introduction -- 7.2 Neural Networks -- 7.2.1 Basics of Neural Networks -- 7.2.2 Training of Neural Networks -- 7.2.3 Examples of Neural Networks -- 7.3 Convolutional Neural Network (CNN) -- 7.3.1 Basics of CNN -- 7.3.2 Learning in CNN -- 7.3.3 Testing in CNN -- 7.3.4 Example of CNN -- 7.4 Pre-trained Models -- 7.4.1 Basics of Pre-trained Models -- 7.4.2 Example of Pre-trained Models -- 7.5 Transfer Learning -- 7.5.1 Basics of Transfer Learning -- 7.5.2 Training in Transfer Learning -- 7.5.3 Example of Transfer Learning -- 7.6 Generative Adversarial Networks (GANs) -- 7.6.1 Basics of GAN -- 7.6.2 Training of GAN -- 7.6.3 Implementation of GAN -- 7.6.4 Example of GAN -- 7.7 Detection Methods -- 7.7.1 Basics of Object Detection -- 7.7.2 Region Based Methods -- 7.7.3 YOLO -- 7.7.4 SSD -- 7.7.5 RetinaNet -- 7.7.6 Examples of Object Detection -- 7.8 Summary -- References -- 8 Simulation in X-ray Testing -- 8.1 Introduction -- 8.2 Modeling -- 8.2.1 Geometric Model -- 8.2.2 X-ray Imaging -- 8.3 Basic General Simulation -- 8.4 Flaw Simulation -- 8.4.1 Mask Superimposition -- 8.4.2 CAD Models for Object and Defect -- 8.4.3 CAD Models for Defects Only -- 8.5 Superimposition Using Multiplication of Images
  • 8.6 Simulation of X-ray Images Using GAN -- 8.7 Simulation with aRTist -- 8.8 Summary -- References -- 9 Applications in X-ray Testing -- 9.1 Introduction -- 9.2 Castings -- 9.2.1 State of the Art -- 9.2.2 An Application -- 9.2.3 An Example -- 9.3 Welds -- 9.3.1 State of the Art -- 9.3.2 An Application -- 9.3.3 An Example -- 9.4 Baggage -- 9.4.1 State of the Art -- 9.4.2 An Application -- 9.4.3 An Example Using Multiple Views -- 9.4.4 Example Using Deep Learning -- 9.5 Natural Products -- 9.5.1 State of the Art -- 9.5.2 An Application -- 9.5.3 An Example -- 9.6 Further Applications -- 9.6.1 Cargo Inspection -- 9.6.2 Electronic Circuits -- 9.7 Summary -- References -- Appendix A mathbbGDXray+ Database -- Index
  • 3.5.3 Correspondence Between Four Views or More -- 3.6 Three-Dimensional Reconstruction -- 3.6.1 Linear 3D Reconstruction from Two Views -- 3.6.2 3D Reconstruction from Two or More Views -- 3.7 Summary -- References -- 4 X-Ray Image Processing -- 4.1 Introduction -- 4.2 Image Preprocessing -- 4.2.1 Noise Removal -- 4.2.2 Contrast Enhancement -- 4.2.3 Shading Correction -- 4.3 Image Filtering -- 4.3.1 Linear Filtering -- 4.3.2 Non-linear Filtering -- 4.4 Edge Detection -- 4.4.1 Gradient Estimation -- 4.4.2 Laplacian-of-Gaussian (LoG) -- 4.4.3 Canny Edge Detector -- 4.5 Segmentation -- 4.5.1 Thresholding -- 4.5.2 Region Growing -- 4.5.3 Maximally Stable Extremal Regions (MSER) -- 4.6 Image Restoration -- 4.7 Summary -- References -- 5 X-ray Image Representation -- 5.1 Introduction -- 5.2 Geometric Features -- 5.2.1 Basic Geometric Features -- 5.2.2 Elliptical Features -- 5.2.3 Fourier Descriptors -- 5.2.4 Invariant Moments -- 5.3 Intensity Features -- 5.3.1 Basic Intensity Features -- 5.3.2 Contrast -- 5.3.3 Crossing Line Profiles -- 5.3.4 Intensity Moments -- 5.3.5 Statistical Textures -- 5.3.6 Gabor -- 5.3.7 Filter Banks -- 5.4 Descriptors -- 5.4.1 Local Binary Patterns -- 5.4.2 Binarized Statistical Image Features (BSIF) -- 5.4.3 Histogram of Oriented Gradients -- 5.4.4 Scale-Invariant Feature Transform (SIFT) -- 5.5 Sparse Representations -- 5.5.1 Traditional Dictionaries -- 5.5.2 Sparse Dictionaries -- 5.5.3 Dictionary Learning -- 5.6 Feature Selection -- 5.6.1 Basics -- 5.6.2 Exhaustive Search -- 5.6.3 Branch and Bound -- 5.6.4 Sequential Forward Selection -- 5.6.5 Sequential Backward Selection -- 5.6.6 Ranking by Class Separability Criteria -- 5.6.7 Forward Orthogonal Search -- 5.6.8 Least Square Estimation -- 5.6.9 Combination with Principal Components -- 5.6.10 Feature Selection Based in Mutual Information -- 5.7 A Final Example -- 5.8 Summary