book

Learning OpenCV 3 Computer Vision with Python (Update)

by Joe Minichino, Joseph Howse

September 2015

Intermediate to advanced

266 pages

5h 38m

English

Packt Publishing

Read now

Unlock full access

Support files, eBooks, discount offers, and moreWhy subscribe?Free access for Packt account holders
What this book covers

Downloading the example codeErrataPiracyQuestions
Choosing and using the right setup toolsInstallation on WindowsUsing binary installers (no support for depth cameras)Using CMake and compilersInstalling on OS XUsing MacPorts with ready-made packagesUsing MacPorts with your own custom packagesUsing Homebrew with ready-made packages (no support for depth cameras)Using Homebrew with your own custom packagesInstallation on Ubuntu and its derivativesUsing the Ubuntu repository (no support for depth cameras)Building OpenCV from a sourceInstallation on other Unix-like systems
Basic I/O scriptsReading/writing an image fileConverting between an image and raw bytesAccessing image data with numpy.arrayReading/writing a video fileCapturing camera framesDisplaying images in a windowDisplaying camera frames in a window
Abstracting a video stream with managers.CaptureManagerAbstracting a window and keyboard with managers.WindowManagerApplying everything with cameo.Cameo
Converting between different color spacesA quick note on BGR
High pass filterLow pass filter
Line detectionCircle detection
Creating modules
Example of foreground detection with GrabCutImage segmentation with the Watershed algorithm
Conceptualizing Haar cascades
Performing face detection on a still imagePerforming face detection on a videoPerforming face recognitionGenerating the data for face recognitionRecognizing facesPreparing the training dataLoading the data and recognizing facesPerforming an Eigenfaces recognitionPerforming face recognition with FisherfacesPerforming face recognition with LBPHDiscarding results with confidence score
Feature detection algorithmsDefining featuresDetecting features – cornersFeature extraction and description using DoG and SIFTAnatomy of a keypointFeature extraction and detection using Fast Hessian and SURFORB feature detection and feature matchingFASTBRIEFBrute-Force matchingFeature matching with ORBUsing K-Nearest Neighbors matchingFLANN-based matchingFLANN matching with homographyA sample application – tattoo forensicsSaving image descriptors to fileScanning for matches
Object detection and recognition techniquesHOG descriptorsThe scale issueThe location issueImage pyramidSliding windowsNon-maximum (or non-maxima) suppressionSupport vector machinesPeople detectionCreating and training an object detectorBag-of-wordsBOW in computer visionThe k-means clustering
What did we just do?SVM and sliding windowsExample – car detection in a sceneExamining detector.pyAssociating training data with classesDude, where's my car?
Detecting moving objectsBasic motion detection
Meanshift and CAMShiftColor histogramsThe calcHist functionThe calcBackProject functionIn summaryBack to the code
Predict and updateAn exampleA real-life example – tracking pedestriansThe application workflowA brief digression – functional versus object-oriented programmingThe Pedestrian classThe main programWhere do we go from here?
Artificial neural networksNeurons and perceptrons
Network layers by exampleThe input layerThe output layerThe hidden layerThe learning algorithms
ANN-imal classificationTraining epochs
MNIST – the handwritten digit databaseCustomized training dataThe initial parametersThe input layerThe hidden layerThe output layerTraining epochsOther parametersMini-librariesThe main file
ImprovementsPotential applications
To boldly go…

Content preview from Learning OpenCV 3 Computer Vision with Python (Update)

Getting Haar cascade data

Once you have a copy of the source code of OpenCV 3, you will find a folder, data/haarcascades.

This folder contains all the XML files used by the OpenCV face detection engine to detect faces in still images, videos, and camera feeds.

Once you find haarcascades, create a directory for your project; in this folder, create a subfolder called cascades, and copy the following files from haarcascades into cascades:

haarcascade_profileface.xml
haarcascade_righteye_2splits.xml
haarcascade_russian_plate_number.xml
haarcascade_smile.xml
haarcascade_upperbody.xml

As their names suggest, these cascades are for tracking faces, eyes, noses, and mouths. They require a frontal, upright view of the subject. We will use them later when building ...