Scribd
Upload
45 views

Intro To The Opencv Library: For Tu Dresden Computer Vision 2 Lecture and General Use

The document introduces the OpenCV library, including why it was created, what it is capable of, how to install it, and examples of code snippets. It provides an overview of the library's history and development. Key topics covered include the library's extensive algorithm and function collection, cross-platform capabilities, real-time performance, and liberal license. Installation instructions and links to documentation and support forums are also outlined. Users are tasked with setting up their environment and coding a simple image transformation program using OpenCV.

Uploaded by

Hiro
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
45 views

Intro To The Opencv Library: For Tu Dresden Computer Vision 2 Lecture and General Use

The document introduces the OpenCV library, including why it was created, what it is capable of, how to install it, and examples of code snippets. It provides an overview of the library's history and development. Key topics covered include the library's extensive algorithm and function collection, cross-platform capabilities, real-time performance, and liberal license. Installation instructions and links to documentation and support forums are also outlined. Users are tasked with setting up their environment and coding a simple image transformation program using OpenCV.

Uploaded by

Hiro
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 16

Intro to the OpenCV Library

for TU Dresden Computer Vision 2


lecture and general use
(some slides are cc from 'opencv 3.0' Kirill Kornyakov, Itseez)
Topics

1.Why
2.What
3.Install
4.Example Project
5.Your Task
6.Your Questions → in Opal forum

The OpenCV Library


Why OpenCV?

1. 2,500+ algorithms and functions


2. Cross-platform, portable API
3. Real-time performance
4. Liberal BSD license
5. fast and regular updates

The OpenCV Library


The OpenCV Library
History

2001: Intel → open src, 2008: Willow Garage, itSeez, 2010: Nvidia,
2016: Intel buys itSeez, Machine Learning rise (people move from cv to ml)

The OpenCV Library


What? core module tutorials:
Mat - The Basic Image Container

How to scan images, lookup tables and time measurement with OpenCV

Mask operations on matrices

Adding (blending) two images using OpenCV

Changing the contrast and brightness of an image

Basic Drawing

Random generator and text with OpenCV

Discrete Fourier Transform

File Input and Output using XML and YAML files

The OpenCV Library


What? imgproc module tutorials:
Smoothing Images Histogram Equalization
Eroding and Dilating Histogram Calculation
More Morphology Transformations Histogram Comparison
Image Pyramids Back Projection
Basic Thresholding Operations Template Matching
Making your own linear filters! Finding contours in your image
Adding borders to your images Convex Hull
Sobel Derivatives Creating Bounding boxes
Laplace Operator and circles for contours
Canny Edge Detector Creating Bounding rotated boxes
Hough Line Transform and ellipses for contours
Hough Circle Transform Image Moments
Remapping Point Polygon Test
Affine Transformations

The OpenCV Library


What? Other modules:
Highgui: feature2d:
Adding a Trackbar to our Harris corner detector
applications! Shi-Tomasi corner detector
Video Input with OpenCV Creating your own corner
Creating a video with OpenCV detector and
similarity measurement
calib3d: Detecting corners location in
Camera calibration subpixeles
Feature Description
ml: Feature Matching with FLANN
Support Vector Machines Features2D + Homography to
Deep Neural Networks (dnn) find a known object
Principal Component Analysis Detection of planar objects

objdetect:
Cascade Classifier

The OpenCV Library


What? code examples:
Hello World example shows an image:

int main(int argc, char** argv)


{
Mat img = imread(argv[1], 1);
imshow("", img);
waitKey();
return 0;
}

The OpenCV Library


What? code examples:
int main(int argc, char** argv)
{
Mat img, gray;
img = imread(argv[1], 1);
imshow("original", img);
cvtColor(img, gray, COLOR_BGR2GRAY);
GaussianBlur(gray, gray, Size(7, 7),
1.5);
Canny(gray, gray, 0, 50);
imshow("edges", gray);
waitKey();
return 0;
}

The OpenCV Library


What? code examples:
Threshold:
Mat emptyPixImg = GrayImg < 1;

Image from (Camera- or) Directory-stream:


VideoCapture cap("TextureImages/Texture_%02d_inpaint.png");
Mat Img;
cap >> Img;

Create a 2D-Gaussian:
Mat Gauss2D = Mat::zeros(TemplateWidth, TemplateWidth,
CV_32FC1);
Gauss2D.at<float>( TemplateHW, TemplateHW) = 1.0;
GaussianBlur(Gauss2D, Gauss2D, Size(TemplateWidth,
TemplateWidth), sigma, sigma);

The OpenCV Library


What? code examples:
pointer work to speed up inner loops (useful in debug mode):

(1)
int** iim = new int*[h];
for (y=0; y<h; y++)
{ iim[y] = IntegralImg.ptr<int>(y);
}
int diffy = 2*( iim[y][x+dx] - iim[y][x-dx] ) +
iim[y-dy][x-dx] - iim[y-dy][x+dx] +
iim[y+dy][x-dx] – iim[y+dy][x+dx];

(2)
float *pCR, *pCRData = (float*) CorrResult.data;
*pCR = pCRData + y*w;
for ( int x = TemplateWidth; x < w-TemplateWidth; x++ )
{ pCR[x] = ssd; // write ssd result to result image
}

The OpenCV Library


How?

1. Home: opencv.org
2. Documentation: docs.opencv.org
3. Q&A forum: answers.opencv.org
4. Report issues: github.com/opencv/opencv/issues
5. Develop: github.com/opencv/opencv

The OpenCV Library


How? Install:
1. download:https://github.com/opencv/opencv/
2. run Cmake(gui), check/install add-ons and
configure until all problems have gone
create debug libs from source code, include Qt
generate makefile or .sln
3. make / compile
4. sudo make install (depending on environment)
(I use win+visual studio only currently, otherwise debian and kdevelop,
QtCreator may also be a choice if you do not already have your environment)

5. setup your IDE for OpenCV (set path, includes ...)


6. run example(s)
Thus you can run AND debug into any OpenCV
example → good to read the code and for debugging

The OpenCV Library


What YOU do::

1. download OpenCV source code from:


https://github.com/opencv/opencv/
2. Install additionally Libs (Qt, Eigen, …) and
compile OpenCV; create debug libs!
3. Setup IDE and run edges example

You have one week for this task.

The OpenCV Library


What YOU do:

1.Set up your development environment and


make a (simple ?) grey value transformation program.
2. You are free to use opencv and other example code you find,
3. but should
- put it all together on your own
- cite your source in a comment.
4.Good C++ coding style and a lot of comments!
5.Your code must compile without errors on Win
and Linux systems (i.e. avoid Win-specific code).

Here is a link to the pdf for the details of coding task.


You have another week for this coding task,
i.e. until 2020-04-20 (maybe more because of Easter).

The OpenCV Library

You might also like