openCV with python
- 1.
- 2. Installations - windows •Python 2.7 or 3.5 https://www.python.org/downloads/ • Frequently used Libraries : Numpy, Matplotlib, Scipy http://www.scipy.org/ • Or try Anaconda (Python + popular libraries) https://www.continuum.io/downloads • openCV 3.0.0 http://opencv.org/downloads.html • Copy “cv2.pyd” from pathopencvbuildpython2.7x86(x64) Paste “cv2.pyd” to pathAnacondaLibsite-packages In python shell : import cv2 cv2.__version__ (two underscores) ‘3.0.0’
- 3. Installations - mac •Python 2.7 or 3.5 or Anaconda https://www.python.org/downloads/ https://www.continuum.io/downloads • Frequently used Libraries : Numpy, Matplotlib, Scipy in Terminals : pip install –U numpy scipy matplotlib (for Python 2.X) pip3 install –U numpy scipy matplotlib (for Python 3.X) • Download openCV http://opencv.org/downloads.html • Install Xcode https://developer.apple.com/xcode/ • Install Cmake https://cmake.org/download/ • Build openCV for Python http://luigolas.com/blog/2014/09/15/install-opencv3-with- python-3-mac-osx/
- 4. Ex1: read/show images •Put the image (lena.bmp) and cv_ex1.py in the same folder •
- 5.
- 6. Ex2: open videos •Put the video (test.mp4) and cv_ex2.py in the same folder •
- 7.
- 8. Ex3: using webCam •Very similar to video open, just change the argument to be 0 •
- 9. Ex4: edges detection •Adopt “canny” method to achieve edges detection •
- 10.
- 11. Ex5: lines detection •Lines detection using Hough Transform line equation using (r, thita) •
- 12.
- 13. Ex6: object tracking usingcolor filtering • using color as features : quick and dirty • Detecting something in blue
- 14. • Ex6: object tracking usingcolor filtering
- 15. Ex7: contours classification •blue objects detextion shapes recognition • Rectangle shown in green • multiple-sides-polygon (like circles) shown in red
- 16.
- 17. Lightening condition invariant? •Color looks different with lightening issue.
- 18. Finding corners’ features •Color invariant, Scale invariant
- 19. Finding corners’ features •Color invariant, Scale invariant
- 20. Features matching • featuresmatching with two images
- 21. • Features method: ORB (Oriented FAST and Rotated BRIEF) • Matching method : BruteForceMatcher • Ex8: object tracking using corners’ features
- 22. • Ex8: object tracking usingcorners’ features
- 23. Ex9: Face recognition/tracking •Download the training model (haarcascade_frontalface_default.xml') and put it in the same folder with the code cv_ex9.py. •
- 24.
- 25. References • You canaccess the example from https://github.com/ntudavid/CV_lab • Reference : OpenCV – Python Tutorials https://opencv-python-tutroals.readthedocs.org/en/latest/