Design and develop advanced computer vision projects using OpenCV with PythonAbout This BookProgram advanced computer vision applications in Python using different features of the OpenCV libraryPractical end-to-end project covering an important computer vision problemAll projects in the book include a step-by-step guide to create computer vision applicationsWho This Book Is For This book is for intermediate users of OpenCV who aim to master their skills by developing advanced practical applications. Readers are expected to be familiar with OpenCV's concepts and Python libraries. Basic knowledge of Python programming is expected and assumed.What You Will LearnGenerate real-time visual effects using different filters and image manipulation techniques such as dodging and burningRecognize hand gestures in real time and perform hand-shape analysis based on the output of a Microsoft Kinect sensorLearn feature extraction and feature matching for tracking arbitrary objects of interestReconstruct a 3D real-world scene from 2D camera motion and common camera reprojection techniquesTrack visually salient objects by searching for and focusing on important regions of an imageDetect faces using a cascade classifier and recognize emotional expressions in human faces using multi-layer peceptrons (MLPs)Recognize street signs using a multi-class adaptation of support vector machines (SVMs)Strengthen your OpenCV2 skills and learn how to use new OpenCV3 featuresIn Detail OpenCV is a native cross platform C++ Library for computer vision, machine learning, and image processing. It is increasingly being adopted in Python for development. OpenCV has C++/C, Python, and Java interfaces with support for Windows, Linux, Mac, iOS, and Android. Developers using OpenCV build applications to process visual data; this can include live streaming data from a device like a camera, such as photographs or videos. OpenCV offers extensive libraries with over 500 functions This book demonstrates how to develop a series of intermediate to advanced projects using OpenCV and Python, rather than teaching the core concepts of OpenCV in theoretical lessons. Instead, the working projects developed in this book teach the reader how to apply their theoretical knowledge to topics such as image manipulation, augmented reality, object tracking, 3D scene reconstruction, statistical learning, and object categorization. By the end of this book, readers will be OpenCV experts whose newly gained experience allows them to develop their own advanced computer vision applications.Style and approach This book covers independent hands-on projects that teach important computer vision concepts like image processing and machine learning for OpenCV with multiple examples. Cover......Page 1 Copyright......Page 3 Credits......Page 4 About the Author......Page 5 About the Reviewers......Page 6 www.PacktPub.com......Page 8 Table of Contents......Page 10 Preface......Page 16 Chapter 1: Fun with Filters......Page 22 Creating a black-and-white pencil sketch......Page 24 Implementing dodging and burning in OpenCV......Page 25 Pencil sketch transformation......Page 27 Generating a warming/cooling filter......Page 29 Implementing a curve filter by using lookup tables......Page 30 Designing the warming/cooling effect......Page 31 Using a bilateral filter for edge-aware smoothing......Page 34 Detecting and emphasizing prominent edges......Page 35 Combining colors and outlines to produce a cartoon......Page 37 Running the app......Page 38 The GUI base class......Page 39 Handling video streams......Page 41 A basic GUI layout......Page 42 A custom filter layout......Page 43 Summary......Page 45 Chapter 2: Hand Gesture Recognition Using a Kinect Depth Sensor......Page 46 Setting up the app......Page 48 Accessing the Kinect 3D sensor......Page 49 Running the app......Page 50 The Kinect GUI......Page 51 Tracking hand gestures in real time......Page 52 Hand region segmentation......Page 53 Finding the most prominent depth of the image center region......Page 54 Applying morphological closing to smoothen the segmentation mask......Page 56 Finding connected components in a segmentation mask......Page 57 Determining the contour of the segmented hand region......Page 59 Finding the convex hull of a contour area......Page 60 Hand gesture recognition......Page 61 Distinguishing between different causes of convexity defects......Page 62 Classifying hand gestures based on the number of extended fingers......Page 63 Summary......Page 66 Chapter 3: Finding Objects via Feature Matching and Perspective Transforms......Page 68 Tasks performed by the app......Page 69 Planning the app......Page 71 Running the app......Page 72 The FeatureMatching GUI......Page 73 The process flow......Page 74 Feature detection......Page 76 Detecting features in an image with SURF......Page 77 Feature matching......Page 78 The ratio test for outlier removal......Page 79 Visualizing feature matches......Page 80 Homography estimation......Page 82 Warping the image......Page 85 Feature tracking......Page 86 Early outlier detection and rejection......Page 87 Seeing the algorithm in action......Page 89 Summary......Page 91 Chapter 4: 3D Scene Reconstruction Using Structure from Motion......Page 92 Planning the app......Page 94 Camera calibration......Page 95 The pinhole camera model......Page 96 Estimating the intrinsic camera parameters......Page 97 The camera calibration GUI......Page 98 Initializing the algorithm......Page 99 Collecting image and object points......Page 100 Finding the camera matrix......Page 102 Setting up the app......Page 103 The main function routine......Page 104 The SceneReconstruction3D class......Page 105 Estimating the camera motion from a pair of images......Page 106 Point matching using rich feature descriptors......Page 107 Point matching using optic flow......Page 108 Finding the camera matrices......Page 112 Image rectification......Page 114 Reconstructing the scene......Page 116 3D point cloud visualization......Page 118 Summary......Page 121 Chapter 5: Tracking Visually Salient Objects......Page 122 Setting up the app......Page 125 The main function routine......Page 126 The MultiObjectTracker class......Page 127 Visual saliency......Page 128 Fourier analysis......Page 130 Natural scene statistics......Page 133 Generating a Saliency map with the spectral residual approach......Page 136 Detecting proto-objects in a scene......Page 140 Mean-shift tracking......Page 141 Automatically tracking all players on a soccer field......Page 143 Setting up the necessary bookkeeping for mean-shift tracking......Page 145 Tracking objects with the mean-shift algorithm......Page 146 Putting it all together......Page 148 Summary......Page 149 Chapter 6: Learning to Recognize Traffic Signs......Page 150 Supervised learning......Page 152 The training procedure......Page 153 The testing procedure......Page 154 A classifier base class......Page 156 The GTSRB dataset......Page 157 Parsing the dataset......Page 158 Feature extraction......Page 161 Common preprocessing......Page 162 Speeded Up Robust Features......Page 163 Histogram of Oriented Gradients......Page 164 Support Vector Machine......Page 166 Using SVMs for Multi-class classification......Page 167 Training the SVM......Page 169 Testing the SVM......Page 170 Confusion matrix......Page 173 Accuracy......Page 174 Precision......Page 175 Recall......Page 177 Putting it all together......Page 178 Summary......Page 182 Chapter 7: Learning to Recognize Emotion in Faces......Page 184 Planning the app......Page 186 Haar-based cascade classifiers......Page 188 Using a pre-trained cascade classifier......Page 189 The FaceDetector class......Page 190 Detecting faces in grayscale images......Page 191 Preprocessing detected faces......Page 192 Facial expression recognition......Page 195 Running the screen capture......Page 196 The GUI constructor......Page 197 The GUI layout......Page 199 Processing the current frame......Page 201 Adding a training sample to the training set......Page 202 Dumping the complete training set to file......Page 203 Preprocessing the dataset......Page 204 Principal component analysis......Page 206 Multi-layer perceptrons (MLPs)......Page 208 The perceptron......Page 209 Deep architectures......Page 210 An MLP for facial expression recognition......Page 213 Training the MLP......Page 214 Testing the MLP......Page 215 Running the script......Page 216 Putting it all together......Page 218 Summary......Page 220 Index......Page 222
Design and develop advanced computer vision projects using OpenCV with Python
About This Book
- Program advanced computer vision applications in Python using different features of the OpenCV library
- Practical end-to-end project covering an important computer vision problem
- All projects in the book include a step-by-step guide to create computer vision applications
Who This Book Is For
This book is for intermediate users of OpenCV who aim to master their skills by developing advanced practical applications. Readers are expected to be familiar with OpenCV's concepts and Python libraries. Basic knowledge of Python programming is expected and assumed.
What You Will Learn
- Generate real-time visual effects using different filters and image manipulation techniques such as dodging and burning
- Recognize hand gestures in real time and perform hand-shape analysis based on the output of a Microsoft Kinect sensor
- Learn feature extraction and feature matching for tracking arbitrary objects of interest
- Reconstruct a 3D real-world scene from 2D camera motion and common camera reprojection techniques
- Track visually salient objects by searching for and focusing on important regions of an image
- Detect faces using a cascade classifier and recognize emotional expressions in human faces using multi-layer peceptrons (MLPs)
- Recognize street signs using a multi-class adaptation of support vector machines (SVMs)
- Strengthen your OpenCV2 skills and learn how to use new OpenCV3 features
In Detail
OpenCV is a native cross platform C++ Library for computer vision, machine learning, and image processing. It is increasingly being adopted in Python for development. OpenCV has C++/C, Python, and Java interfaces with support for Windows, Linux, Mac, iOS, and Android. Developers using OpenCV build applications to process visual data; this can include live streaming data from a device like a camera, such as photographs or videos. OpenCV offers extensive libraries with over 500 functions
This book demonstrates how to develop a series of intermediate to advanced projects using OpenCV and Python, rather than teaching the core concepts of OpenCV in theoretical lessons. Instead, the working projects developed in this book teach the reader how to apply their theoretical knowledge to topics such as image manipulation, augmented reality, object tracking, 3D scene reconstruction, statistical learning, and object categorization.
By the end of this book, readers will be OpenCV experts whose newly gained experience allows them to develop their own advanced computer vision applications.
Style and approach
This book covers independent hands-on projects that teach important computer vision concepts like image processing and machine learning for OpenCV with multiple examples.
Design and develop advanced computer vision projects using OpenCV with Python About This Book Who This Book Is For This book is for intermediate users of OpenCV who aim to master their skills by developing advanced practical applications. Readers are expected to be familiar with OpenCV's concepts and Python libraries. Basic knowledge of Python programming is expected and assumed. What You Will Learn In Detail OpenCV is a native cross platform C++ Library for computer vision, machine learning, and image processing. It is increasingly being adopted in Python for development. OpenCV has C++/C, Python, and Java interfaces with support for Windows, Linux, Mac, iOS, and Android. Developers using OpenCV build applications to process visual data; this can include live streaming data from a device like a camera, such as photographs or videos. OpenCV offers extensive libraries with over 500 functions This book demonstrates how to develop a series of intermediate to advanced projects using OpenCV and Python, rather than teaching the core concepts of OpenCV in theoretical lessons. Instead, the working projects developed in this book teach the reader how to apply their theoretical knowledge to topics such as image manipulation, augmented reality, object tracking, 3D scene reconstruction, statistical learning, and object categorization. By the end of this book, readers will be OpenCV experts whose newly gained experience allows them to develop their own advanced computer vision applications. All source code is available on (http://github.com/mbeyeler/opencv-python-blueprints) github.com/mbeyeler/opencv-python-blu... . Style and approach This book covers independent hands-on projects that teach important computer vision concepts like image processing and machine learning for OpenCV with multiple examples. Key FeaturesBook DescriptionWhat you will learn[•]Generate real-time visual effects using different filters and image manipulation techniques such as dodging and burning[•]Recognize hand gestures in real time and perform hand-shape analysis based on the output of a Microsoft Kinect sensor[•]Learn feature extraction and feature matching for tracking arbitrary objects of interest[•]Reconstruct a 3D real-world scene from 2D camera motion and common camera reprojection techniques[•]Track visually salient objects by searching for and focusing on important regions of an image[•]Detect faces using a cascade classifier and recognize emotional expressions in human faces using multi-layer peceptrons (MLPs)[•]Recognize street signs using a multi-class adaptation of support vector machines (SVMs)[•]Strengthen your OpenCV2 skills and learn how to use new OpenCV3 featuresWho this book is forThis book is for intermediate users of OpenCV who aim to master their skills by developing advanced practical applications. Readers are expected to be familiar with OpenCV&apos
