This tutorial is for the people who are complete beginners in the area of computer vision and image processing also aren't very familiar with numpy and other tools in python. It intends to give a good balance between the practical and the theoretical aspects of the area. I expect to finish the tutorial within 2 hours.
(10 minutes)
- The matrix view
- The samples of a function view
- Color spaces
- Anaconda Package for the latest Python 2.x . That should include the python interpretor, IPython, numpy, scipy and matplotlib. Anaconda can be downloaded from Continuum's website. Download Link
- Python imaging library(PIL). (NOTE: PIL is not supported for Python 3.x)
PIL can be installed with
conda
(installed with anaconda) using commandconda install pil
- Very basic familiarity with Python: That can be gained from spending an hour or two on code academy.
(20 minutes)
- Indexing
- Slicing
- Point wise operations
- Operations on regions
- Exercise 1:
(40 minutes)
- Loading an image
- Displaying an image
- Manipulating Pixels
- More manipulation of Pixels
- Exercise 2: The red, blue and green image from the given colored image (teaches about the color space in the RBG image and manipulating them)
- Exercise 3: Write an H on the image (teaches similar things as above)
- Homework 1: Staining the image
(20 minutes)
- Basic 2D plotting with Matplotlib, (ploting sin(x))
- Exercise 4: plot
cos(x)*cos(x)
- Parametric Plotting: plotting a parabola with axis parallel to x axis
- Exercise 5: plot a circle
- What are histograms
- Plotting a histogram
- Exercise: plotting a histogram of colored image
- Linear stretching of histogram
- Histogram equalization
- Homework 2:
(30 minutes)