ADVANCES IN PATTERN RECOGNITION AND IMAGE PROCESSING
Chief Editor: Mengxin Li
Associate Editors: Nan Hu, Jing Hou, Ying Zhang
Shenyang Jianzhu University
Copyright © 2019 by Cayley Nielson Press, Inc.
ISBN: 978-1-7923-2700-1
Cayley Nielson Press Scholarly Monograph Series Book Code No.: 209-7-4
US$232.60
Preface
With the advance of technologies, image recognition and processing has become increasingly widely used in various fields such as optical character recognition, fingerprint, facial recognition and scene recognition. Additionally, the combination of big data and the structuring of image data is also the frontier research direction. This monograph focus on exploring some aspects of the image recognition and image processing.
Visual defect inspection is an important part of quality assurance in many fields of production. A methodology has been proposed to incorporate contemporary techniques of feature classification for visual inspection of surface defects. A rough sets based neural network with fuzzy input (RNNFI) for pattern recognition has been proposed to incorporate the techniques in order to achieve further improvements in performance. The hybrid method for the RNNFI has taken the advantages of all the techniques incorporated. Experiments have been carried out on defect classification of wood veneers.
In order to understand and master novel methods in the field of pattern recognition, some reviews and research are introduced including gait recognition, fire recognition, bare footprint recognition, vehicle detection, moving target inspection. Image processing is also studied covering image inpainting method and 3D reconstruction method.
Much of the book is the result of close work within the development team, which has contributed several passages, answered numerous technical questions, reviewed the manual, and made many corrections of Chapters.
The contributors are:
- Ying Zhang (Chapter 1、2)
- Jing Hou (Chapter 3)
- Jingke Xu (Chapter 3、4)
- Ke Xu (Chapter 4、5)
- Yang Cao (Chapter 5)
- Nan Hu (Chapter 3、5)
- Rui Zhang (Chapter 6)
- Changtao Wang (Chapter 6、7)
- Ziyang Han (Chapter 4、8)
- Jiali Hao (Chapter 8)
- Tianhui Zhang (Chapter 9、16)
- Rui Xu (Chapter 9)
- Jiang Tongwei(Chapter 10、11)
- Feng Jin (Chapter11、12)
- Li Chen (Chapter 13)
- Weijing Xu (Chapter 14、15)
- Shengbin Zhao (Chapter 16)
Mengxin Li
Shenyang Jianzhu University
Shenyang,Liaoning, China
November 10,2019
Contents
Preface.......................................................................................................... 1
Chapter 1 Rough Sets Based Neural Network Pattern Recognition...................... 9
1.1 Introduction............................................................................................... 9
1.2 Variable Precision Rough Sets..................................................................... 9
1.3 Reduction of Condition Attributes with VPRS............................................... 13
1.4 Generation of Probaalilistic Decision Rules................................................. 14
1.5 VPRS for Featrue Classification of Wood Veneer Defects............................... 15
1.6 Rough Sets Based Neural Network Classifier............................................... 19
1.7 Experiments............................................................................................. 21
Chapter 2 A classifier using rough sets based neural network with fuzzy input (RNNFI) 24
2.1 Introduction............................................................................................. 24
2.2 Why Fuzzy Data Processing....................................................................... 25
2.2.1 Fuzzifier............................................................................................... 25
2.2.2 Fuzzy Rule Base.................................................................................... 26
2.2.3 Fuzzy Inference Engine........................................................................... 27
2.2.4 Defuzzifier............................................................................................ 27
2.3 Fuzzy Sets and Rough Sets......................................................................... 28
2.4 A Neural Network with Fuzzy Input for Wood Veneer Inspection..................... 29
2.5 A Classifier Using Rough Sets Based Neural Network with Fuzzy Input........... 30
2.6 Experiments............................................................................................. 32
Chapter 3 Review of Fire Detection Technologies Based on Video Image........... 41
3.1 Introduction............................................................................................. 41
3.2 Fire Detection Techniques......................................................................... 42
3.2.1 Fire Flame Detection in Video................................................................. 42
3.2.2 Fire Smoke Detection in Video................................................................. 49
3.3 Conclusions............................................................................................. 54
Chapter 4 Extraction and Determination of Fire Features in Accurate Motion Area 56
4.1 Introduction............................................................................................. 56
4.2 Moving Region Detection.......................................................................... 57
4.2.1 Background Estimation Method............................................................... 57
4.2.2 Improved Background Estimation Method................................................. 59
4.3 Extraction and Detection of the Fire Features.............................................. 61
4.3.1 Extracting the Color of Fire Flame.......................................................... 61
4.3.2 Extracting the Energy of Fire Flame......................................................... 62
4.3.3 Extracting the Flashing Feature of Fire Flame.......................................... 63
4.3.4 Extracting the Contour Feature of Fire Flame........................................... 64
4.4 Experiments and Results Analysis............................................................... 65
4.5 Conclusion.............................................................................................. 69
Chapter 5 A New Fire Detection Method Based on Integrated Features............. 70
5.1 Introduction............................................................................................. 70
5.2 Gaussian Model....................................................................................... 71
5.3 Color Spaces........................................................................................... 73
5.3.1 RGB Color Space.................................................................................. 73
5.3.2 HSI Color Space.................................................................................... 74
5.4 Foreground Object Detection..................................................................... 75
5.4.1 Background Model................................................................................. 75
5.4.2 Foreground Object Detection.................................................................. 77
5.5 Extract Color Features of The Flame and Detection..................................... 78
5.6 Flicker Characteristics of Flame................................................................ 79
5.7 Spatial Wavelet Energy.............................................................................. 80
5.8 Circularity of Flame................................................................................. 81
5.9 Experiments and Results Analysis............................................................... 82
5.10 Conclusion............................................................................................ 86
Chapter 6 Research of Moving Target Detection Technology in Intelligent Video Surveillance System 87
6.1 Introduction............................................................................................. 87
6.2 Static Background.................................................................................... 88
6.2.1 Background subtraction.......................................................................... 88
6.2.2 Optical flow method............................................................................... 91
6.2.3 Frame difference method........................................................................ 93
6.3 Dynamic Background................................................................................ 94
6.3.1 Motion compensation method.................................................................. 94
6.3.2 Motion segmentation method................................................................... 95
6.3.3 Regional integration method................................................................... 95
6.4 Conclusion.............................................................................................. 95
Chapter 7 Application of Non-Parametric Kernel Density Background Modeling Method in Intelligent Video Surveillance System..... 97
7.1 Introduction.. 97
7.2 The Non-parametric Kernel Density Background Modeling........................... 98
7.3 Adaptive Kernel Width Selection................................................................. 99
7.4 The Foreground Segmentation and Adaptive Background Updating.............. 100
7.5 Hadow Suppression................................................................................ 101
7.6 Test results............................................................................................. 102
7.7 Conclusion............................................................................................ 104
Chapter 8 Anti-occlusion Video Target Tracking Based On Double Threshold Judgment 105
8.1 Introduction........................................................................................... 105
8.2 Target tracking based on DSST algorithm.................................................. 107
8.3 Anti-occlusion Video Target Tracking Based on Double Threshold Judgment.. 109
8.4 Experimental results and analysis of simulation.......................................... 112
8.5 Conclusion............................................................................................. 114
Chapter 9 Adaptive Anti-occlusion Moving Object Trackingof Intelligent Video Surveillance System 115
9.1 Introduction........................................................................................... 115
9.2 Meanshift Algorithm................................................................................ 116
9.2.1 Object model description....................................................................... 116
9.2.2 Candidate model description................................................................. 118
9.3 Moving object tracking using Meanshift algorithm...................................... 118
9.4 Improved Meanshift Algorithm using Double Bhattacharrya Coefficients Discrimination 120
9.5 Conclusion............................................................................................ 123
Chapter 10 Research on Body’s Bare Footprint Recognition Algorithm........... 125
10.1 Introduction......................................................................................... 125
10.2 Footprint Images Segmentation.............................................................. 126
10.2.1 Threshold Value Method...................................................................... 126
10.2.2 Edge Detection Method....................................................................... 128
10.3 Footprint Feature Extraction.................................................................. 130
10.4 Personal Homoousia Identify................................................................. 134
10.5 Conclusion........................................................................................... 137
Chapter 11 A Static Gesture Recognition Algorithm Based on DAG-SVMs....... 139
11.1 Introduction.......................................................................................... 139
11.2 Static Gesture Recognition with DAG-SVMs............................................. 140
11.2.1 Two-dimensional depth histogram........................................................ 140
11.2.2 Edge Extraction and Feature Extraction................................................ 142
11.2.3 DAG-SVMs Static Gesture Recognition Algorithm.................................. 143
11.3 Experiments and Results........................................................................ 145
11.4 Conclusion........................................................................................... 147
Chapter 12 Vehicle Detection Based on Multiple Characteristic Information.... 148
12.1 Introduction......................................................................................... 148
12.2 Establish AOH by Vechicle Shadow......................................................... 148
12.2.1 Segmentation of Shadows.................................................................... 149
12.2.2 Identification of Vehicle Bottom Shadow................................................ 150
12.3 Texture Characteristics Extraction of AOH............................................... 150
12.4 Edge Characteristics Extraction of AOH.................................................. 152
12.5 Feature Extraction Based on Symmetrical Characteristics of Vertical Edges. 153
12.6 Fusion of Multiple Characteristics Information......................................... 154
12.7 Experiments and Conclusions................................................................. 156
Chapter 13 A Hand-Eye Calibration Method for Line Structured Light System. 159
13.1 Introduction......................................................................................... 159
13.2 Measuring Principle of The Line Structed Light System............................. 160
13.3 New Hand-Eye System Calibration Method.............................................. 161
13.3.1 Camera Calibration........................................................................... 161
13.3.2 Light Plane Calibration...................................................................... 163
13.4 Calibration Experiment Results.............................................................. 166
13.5 Conclusion........................................................................................... 168
Chapter 14 A New Image Inpainting Method based on TV Model.................... 169
14.1 Introduction......................................................................................... 169
14.2 Related Work........................................................................................ 170
14.3 Some Ideal About TV Model................................................................... 171
14.4 A New Image Inpainting Method Based on Tv Model................................. 173
14.5 A New Image Inpainting Method Based on TV Model................................ 175
14.6 Conclusion........................................................................................... 177
Chapter 15 Overview of 3D Reconstruction Methods Based on Multi-view........ 178
15.1 Introduction......................................................................................... 178
15.2 Main Research Contents........................................................................ 180
15.2.1 Feature Detection and Matching.......................................................... 180
15.2.2 Fundamental Matrix Estimation........................................................... 182
15.2.3 Camera Self-Calibration..................................................................... 184
15.2.4 3D Reconstruction.............................................................................. 186
15.2.5 Dense Surface Reconstruction.............................................................. 187
15.3 Problems and Future Works................................................................... 189
15.4 Conclusion........................................................................................... 190
Chapter 16 A Cluster-based PMVS Algorithm with Geometric Constraint........ 191
16.1 Introduction......................................................................................... 191
16.2 PMVS Algorithm................................................................................... 192
16.2.1 Basic Definition................................................................................. 193
16.2.2 Specific Procedure.............................................................................. 194
16.3 PMVS Algorithm with Geometric Constraint............................................ 196
16.4 The Algorithm Based on Clustering......................................................... 197
16.4.1 View Clustering.................................................................................. 197
16.4.2 MVS Filtering.................................................................................... 199
16.5 Experiments and Analysis...................................................................... 200
16.6 Conclusion........................................................................................... 205
References.................................................................................................. 206
Readership
This book should be useful for students, scientists, engineers and professionals working in the areas of optoelectronic packaging, photonic devices, semiconductor technology, materials science, polymer science, electrical and electronics engineering. This book could be used for one semester course on adhesives for photonics packaging designed for both undergraduate and graduate engineering students.
