Gollapudi Learn Computer Vision Using OpenCV

With Deep Learning CNNs and RNNs
1. Auflage 2019
ISBN: 978-1-4842-4261-2
Verlag: Apress
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 151 Seiten

Build practical applications of computer vision using the OpenCV library with Python. This book discusses different facets of computer vision such as image and object detection, tracking and motion analysis and their applications with examples. 

The author starts with an introduction to computer vision followed by setting up OpenCV from scratch using Python. The next section discusses specialized image processing and segmentation and how images are stored and processed by a computer. This involves pattern recognition and image tagging using the OpenCV library. Next, you’ll work with object detection, video storage and interpretation, and human detection using OpenCV. Tracking and motion is also discussed in detail. The book also discusses creating complex deep learning models with CNN and RNN. The author finally concludes with recent applications and trends in computer vision.

After reading this book, you will be able to understand and implement computer vision and its applications with OpenCV using Python. You will also be able to create deep learning models with CNN and RNN and understand how these cutting-edge deep learning architectures work.

What You Will Learn

• Understand what computer vision is, and its overall application in intelligent automation systems
  
• Discover the deep learning techniques required to build computer vision applications
  
• Build complex computer vision applications using the latest techniques in OpenCV, Python, and NumPy
  
• Create practical applications and implementations such as face detection and recognition, handwriting recognition, object detection, and tracking and motion analysis
  

Who This Book Is For

Those who have a basic understanding of machine learning and Python and are looking to learn computer vision and its applications. 

Autoren/Hrsg.

Gollapudi, Sunila

Inhaltsverzeichnis

1;Table of Contents;5
2;About the Author;9
3;About the Technical Reviewer;10
4;Acknowledgments;11
5;Foreword;12
6;Introduction;13
7;Chapter 1: Artificial Intelligence and Computer Vision;17
7.1;Introduction to Artificial Intelligence;19
7.1.1;Natural Language Processing;23
7.1.2;Robotics;26
7.1.3;Machine Learning;27
7.1.4;Expert Systems;29
7.1.5;Speech and Voice Recognition;29
7.1.6;Intelligent Process Automation;30
7.2;Introduction to Computer Vision;30
7.2.1;Scope;31
7.2.2;Challenges of Computer Vision;35
7.2.3;Real-World Applications of Computer Vision;37
7.2.3.1;Automotive Industry;38
7.2.3.2;Healthcare and Biomedical Industry;38
7.2.3.3;Retail Industry;39
7.2.4;Images and Their Features;40
7.2.4.1;Color Spaces;41
7.2.5;Core Building Blocks (Input – Process – Output);42
7.2.5.1;Optical Character Recognition and Intelligent Character Recognition;44
7.2.5.2;Optical Mark Recognition;44
7.3;Conclusion;44
8;Chapter 2: OpenCV with Python;46
8.1;About OpenCV;47
8.2;Setting Up OpenCV with Python;47
8.2.1;Windows Installation;47
8.2.2;macOS Installation;51
8.3;Using Modules;53
8.4;Working with Images and Videos;55
8.4.1;Using NumPy;55
8.4.1.1;Reading and Loading Images with OpenCV and NumPy;56
8.4.1.2;Working with a Histogram Representation;59
8.4.2;Videos;61
8.4.2.1;Loading Videos from a Webcam;61
8.4.2.2;Loading Videos from a File;62
8.4.2.3;Reading the Video and Writing into a File;63
8.5;Conclusion;64
9;Chapter 3: Deep Learning for Computer Vision;66
9.1;Deep Learning: An Overview;67
9.2;Deep Learning Applications in Computer Vision;68
9.2.1;Classification;68
9.2.2;Detection and Localization;69
9.2.3;(Semantic) Segmentation;70
9.2.4;Similarity Learning;70
9.3;Image Captioning;70
9.4;Generative Models;71
9.5;Video Analysis;72
9.6;Neural Networks at Their Core;72
9.6.1;Artificial Neural Networks;73
9.6.2;Artificial Neurons or Perceptrons;73
9.6.3;Training Neural Networks;77
9.6.3.1;Backpropagation;77
9.6.3.2;Gradient Descent and Stochastic Gradient Descent;78
9.7;Convolutional Neural Networks;78
9.7.1;Convolution Layer;79
9.7.2;Pooling Layer;80
9.7.3;Fully Connected Layer;80
9.8;Recurrent Neural Networks;81
9.8.1;Backpropagation Through Time;83
9.9;Conclusion;84
10;Chapter 4: Image Manipulation and Segmentation;85
10.1;Image Manipulations;86
10.1.1;Accessing and Manipulating Pixels;87
10.1.2;Drawing Geometric Shapes or Writing Text on a Color Image;89
10.1.3;Filtering Images;93
10.1.4;Transforming Images;96
10.1.4.1;Translation;97
10.1.4.2;Rotation;99
10.1.4.3;Image Scaling;101
10.1.4.4;Edge Detection;102
10.2;Image Segmentation;104
10.2.1;Line Detection;106
10.2.2;Circle Detection;107
10.3;Conclusion;110
11;Chapter 5: Object Detection and Recognition;111
11.1;Basics of Object Detection;111
11.1.1;Object Detection vs. Object Recognition;112
11.1.2;Template Matching;113
11.1.3;Challenges with Template Matching;116
11.1.4;Understanding Image “Features”;116
11.1.4.1;Interesting and Uninteresting Points;117
11.1.4.2;Types of Image Features;118
11.2;Feature Matching;119
11.2.1;Image Corners As Features;119
11.2.2;Harris Corner Algorithm;120
11.2.3;Feature Tracking and Matching Flow;122
11.2.4;Scale Variant Feature Transform;123
11.2.5;Speeded-Up Robust Features;126
11.2.6;Features from Accelerated Segment Test;127
11.2.7;Binary Robust Independent Elementary Features;128
11.2.8;Oriented FAST and Rotated BRIEF;130
11.3;Conclusion;131
12;Chapter 6: Motion Analysis and Object Tracking;132
12.1;Introduction to Object Tracking;133
12.2;Challenges of Object Tracking;134
12.3;Object Detection Techniques for Tracking;134
12.3.1;Frame Differentiation;135
12.3.2;Background Subtraction;136
12.3.3;Optical Flow;138
12.3.3.1;Lucas–Kanade Differential Algorithm;139
12.3.3.2;Dense Optical Flow Algorithm;142
12.4;Object Classification;144
12.4.1;Shaped-Based Classification;145
12.4.2;Motion-Based Classification;145
12.4.3;Color-Based Classification;145
12.4.4;Texture-Based Classification;146
12.5;Object Tracking Methods;146
12.5.1;Point Tracking Method;147
12.5.2;Kernel-Based Tracking Methods;148
12.5.2.1;Simple Template Matching;148
12.5.2.2;Meanshift Method;149
12.5.2.3;Support Vector Machine;157
12.5.2.4;Layering-Based Tracking;157
12.5.3;Silhouette-Based Tracking;157
12.5.3.1;Contour Tracking;158
12.5.3.2;Shape Matching;158
12.6;Conclusion;158
13;Index;159