Python for Computer Vision - Revision

Python for Computer Vision

Ahmed Fawzy Gad

ahmed.fawzy@ci.menofia.edu.eg

MENOUFIA UNIVERSITYFACULTY OF COMPUTERS AND INFORMATION

INFORMATION TECHNOLOGYCOMPUTER VISION

جامعة المنوفية

كلية الحاسبات والمعلومات

تكنولوجيا المعلومات

الرؤية بالحاسب

جامعة المنوفية

Tuesday 26 September 2017

Index

• Traditional Data Storage in Python

• NumPy Arrays

• Matplotlib

• SciPy

Goals of Computer Vision

• Computer vision aims to enable computer to see, identify objects,and analyze such images to understand them like or better thanhumans.

• To do this, computer needs to store and process images to get usefulinformation.

CAT

Storing Data in Python

ListTuple

Data Structures

Which one is suitable for storing images?

Let`s See.

List Vs. Tuple

• Image is MD. Which one supports MD storage?

• Which one supports updating its elements?

Both

ListTuples are immutable.

Python List for Image Storage

• List is mutable and thuswe can edit the imagepixels easily and applyoperations.

• So, lets start storing animage into a Python list.The following code readsan image in the img_listvariable.

Very time consuming for simple operations

Very time consuming for simple operations.

• Why not applying arithmetic operations rather than looping?

img_list = img_list + 50

List is complex. What is the alternative?

• List adds more complexity in making operations over the images. Onedrawback was seen previously is that list operations are timeconsuming because they require pixel by pixel processing.

• The best way for storing images is using arrays.

• Rather than being time efficient in processing images, arrays hasmany other advantages. Can you imagine what?

Python Arrays• Lists are already available in Python. To use Python arrays, additional

libraries must be used.

• The library supporting arrays in Python is called Numerical Python(NumPy).

• NumPy can be installed using Python command-line.

• It is also available in all-in-one packages like Anaconda.

Installing NumPy

• Based on your environment, you can install new modules.

• For traditional Python distributions, use the PIP installer.

pip install numpy

• For Anaconda, use the conda installed

conda install numpy

• But it is by default available in Anaconda.

Importing a Python Module

• After installing NumPy, we can import it in our programs and scripts.

NumPy Array for MD Data

Matplotlib: Displaying the Image

Array Data Type

• The problem is expecting uint8 data type but another data type wasused.

• To know what is the array data type, use the dtype array property.

• Change array type to uint8.

How to make the conversion to uint8?

Controlling Array dtype

• When creating the array, set the dtype argument of numpy.array tothe desired data type.

• dtype argument can be set to multiple types.

Creating Array with dtype of uint8

Controlling Array dtype

• After array being created, use the astype method of numpy.ndarray.

• It make a new copy of the array after being casted to the specifiedtype in the dtype argument.

Array Operations

• Arithemetic Operations

• Operations between arrays

More Array Creation Methods

• Array Creation

• arange

• linspace

arange vs. linspace

Array Indexing & Slicing

• Indexing can be forward or backward.

• Forward indexing: from start to end.

• Backward indexing: from end to start.

• General form of indexing:

my_array[start:stop:step]

• In backward indexing, the index of the last element is -1.

Start End0 2

End Start-3 -1

Forward Backward

Indexing & Slicing Examples – 1D Array

• Forward: my_array[start=0:stop=6:step=2]

• Backward: my_array[start=-1:stop=-6:step=-2]

• Get all elements starting from index 3

Indexing & Slicing Examples – 2D Array

• For MD arrays, indexing can be applied for each individual dimension. Intersection between the different dimensions will be returned.

• Forward: my_array[start=0:stop=3:step=2, start=1:stop=4:step=1]

• Forward: my_array[start=-1:stop=-3:step=-1, start=0:stop=3:step=1]

Iterating Through Arrays

For

While

Matplotlib: Plotting Data

Scientific Python (SciPy)

• The main use of NumPy is to support numericalarrays in Python. According to the officialdocumentation, NumPy supports nothing but thearray data type and most basic operations:indexing, sorting, reshaping, basic elementwisefunctions, etc.

• SciPy supports everything in NumPy but also addsnew features not existing in NumPy. We canimagine that NumPy is a subset of SciPy.

• Let`s explore what is in SciPy.

SciPy

NumPy

SciPy

• SciPy contains a collection of algorithms and functions based onNumPy. User can use high-level commands to perform complexoperations.

• SciPy is organized into a number of subpackages.

SciPy for Image Processing• SciPy provides modules for working with images from reading,

processing, and saving an image.

• This example applies the Sobel edge detector to an image using SciPy.

References

• SciPy• https://docs.scipy.org/doc/scipy/reference

• NumPy• https://docs.scipy.org/doc/numpy/reference

• Matplotlib• https://matplotlib.org/contents.html