Scribd
Upload
212 views

OpenCV Tutorial

This document provides an overview and tutorials for OpenCV, an open source computer vision and machine learning software library. The tutorials cover installing OpenCV on different platforms, reading and writing images and video, mouse input and point tracking, and face detection using Haar cascades. The code and documentation for the tutorials are available online and demonstrate important OpenCV concepts like using VideoCapture and VideoWriter objects, the Mat image structure, and functions for image processing, edge detection and more.

Uploaded by

Jaiprasad Reddy
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
212 views

OpenCV Tutorial

This document provides an overview and tutorials for OpenCV, an open source computer vision and machine learning software library. The tutorials cover installing OpenCV on different platforms, reading and writing images and video, mouse input and point tracking, and face detection using Haar cascades. The code and documentation for the tutorials are available online and demonstrate important OpenCV concepts like using VideoCapture and VideoWriter objects, the Mat image structure, and functions for image processing, edge detection and more.

Uploaded by

Jaiprasad Reddy
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 18

OpenCV Tutorial

(OpenCV 2.1)

Denis Aleshin

Overview
1. 2. 3. 4. 5. Installation (Windows/Linux/Mac) Tutorial 1 - Reading/Writing videos Tutorial 2 - Reading/Writing images and image processing Tutorial 3 - Mouse input and point tracking Tutorial 4 - Haar cascades and face detection

Installation
Linux -

sudo apt-get install opencv Mac get macports - http://www.macports.org/ sudo port install opencv OR get XCode - http://developer.apple.com/xcode/ and use the OpenCV framework (1.2 instead of 2.1 unless you build your own). Windows Get visual studio - https://www.dreamspark.com/ Download library - http://sourceforge.net/projects/opencvlibrary/files/opencv-win/2.2/

Making your own projects


To use it, you will likely be writing your code in C++. So you need to link your code to the libraries. Two options: 1 - use an IDE like XCode or Visual Studio Visual Studio - http://opencv.willowgarage.com/wiki/VisualC%2B%2B XCode - http://opencv.willowgarage.com/wiki/Mac_OS_X_OpenCV_Port Other - http://opencv.willowgarage.com/wiki/InstallGuide (See point 4) 2 - create your own makefile Makefile example in tutorials CMake (more advanced) - http://www.cmake.org/

Tutorials
Documentation: OpenCV 2.1 - http://opencv.willowgarage.com/documentation/python/index.html OpenCV 2.0 - http://opencv.willowgarage.com/documentation/index.html These slides will be available online. Code from tutorials will be available online.

Tutorial 1
Video input is handled by the VideoCapture object for files and cameras. link Functions in the tutorial:
VideoCapture capture; // Declares capture. capture.open(0); // Initializes capture from camera. capture.open(inputName); // Initializes capture from file. capture.isOpened(); capture >> frame; // Grabs frame from capture.

Tutorial 1
Important function: capture.get(<flag>); Is used to access capture properties, like size, encoding, color depth, etc... Flags used in tutorial: CV_CAP_PROP_FOURCC CV_CAP_PROP_FPS CV_CAP_PROP_FRAME_WIDTH CV_CAP_PROP_FRAME_HEIGHT More Flags

Tutorial 1
Video output is handled by the VideoWriter object. Link Functions from the tutorial:
VideoWriter videoout; // Declare videoout.open(const string& filename, int fourcc, double fps, Size frameSize, bool isColor=true); videoout.isOpened(); videoout << frame; // Write frame to file. Will probably use the get() function to get these.

Tutorial 1
First bits of UI. Link
cvNamedWindow("video", 1); // Creates a new window. imshow( "video", frame ); // Shows an image in window. char c = waitKey(33); // Waits for key input from user for 33 milliseconds.

Tutorial 1
Nice thing about OpenCV 2.0 and on - storage is managed for you. Images, video captures and writes, windows cleaned up when program exits.

Let's look at the code!

Tutorial 2
Images are stored in the Mat structure. Link Can be used to store any matrix (histograms, voxels, etc...), and built with extra support for images. Pixels stored in adjacent memory (so you can do pointer math for really fast code).
M.create(int height,int width,int depth);

Depth is a CV defined constant. For instance, CV_8UC1 is a single channel 8-bit image (greyscale) CV_32FC3 is a 3-channel 32 bit image (RGB). Allocated and deallocated using reference counting (so no need to worry about it!) See link above for details.

Tutorial 2
Reading and writing images Link
Mat img = imread(String filename); imwrite(String filename, Mat img);

Sanity check:
img.empty();

Image processing.
Huge library! Lots of functions! Link Usual format: Filter(Mat src, Mat dst, options...); Examples from code: cvtColor(out, grey, CV_RGB2GRAY); // Link resize(frame, out, Size(0,0), 1/size, 1/size); // Link Sobel(grey, grey, CV_8UC1, 1, 1); // Link

Tutorial 2
Let's look at the code...

Tutorial 3
Lots of other useful code implemented! Canny edge detector - Link Find circles and lines - Link Cascade Classifiers - Link Histograms and Back-projection - Link Meanshift and Camshift - Link

Tutorial 3
A bit more gui... set a callback for a window:
cvSetMouseCallback(const char* windowName, CvMouseCallback onMouse, void* param=NULL); //Link onMouse(int event, int x, int y, int flags, void* param);

And draw circles (or other stuff) on your image Link


circle(image, Point, radius, color); color is a Scalar object, like Scalar(255,0,0);

Tutorial 3
Tracking... several methods Link
calcOpticalFlowPyrLK(lastframe, frame, lastpoints, points, status, errors);

Let's look at how it does...

Tutorial 4
Haar cascades and code.

You might also like