Computer Vision with Python and OpenCV2 (Part 1 - Basics)

01 September 2019 - 4 mins read time

Contents:

• 1. Matplotlib implementation
• Please note that matplotlib implementation is done in jupyter notebook for this blog post.
  • Load libraries
  • Load image and display using matplotlib’s imshow method
  • Displaying image as COLOR IMAGE
  • Displaying image as BLACK AND WHITE IMAGE
• 2. OpenCV2 implementation
  • Displaying images with CV2’s built-in “imshow” method
  • Running python script from jupyter notebook
    • Output of the script

You can download the jupyter notebook or the project from GitHub here: Github project link

In this blog post let us look into different ways of opening and displaying images using both matplotlib and OpenCV2

Picture used in this tutorial is downloaded from www.pexels.com. I do not own the picture. Picture is available for free download and complete credits goes to Chevanon Photography. Link below.

Credit: Photo by Chevanon Photography from Pexels

PEXELS PHOTO DOWNLOAD LINK

1. Matplotlib implementation

Please note that matplotlib implementation is done in jupyter notebook for this blog post.

Load libraries

# OpenCV2
import cv2

# matplotlib
import matplotlib as mat
import matplotlib.pyplot as plt
%matplotlib inline

print(cv2.__version__)

'4.1.0'

print(mat.__version__)

'3.0.2'

Load image and display using matplotlib’s imshow method

puppies = cv2.imread('puppies.jpg')

plt.figure(figsize = (9,7))
plt.imshow(puppies)

Displaying image as COLOR IMAGE

By default matplotlib and OpenCV’s RGB order is different. This is why we need to convert original image using cv2.COLOR_BGR2RGB method.

For this we use OpenCV2’s COLOR_BGR2RGB method to fix the colors in the picture

color_puppies = cv2.cvtColor(puppies, cv2.COLOR_BGR2RGB)

plt.figure(figsize = (9,7))
plt.imshow(color_puppies)

Displaying image as BLACK AND WHITE IMAGE

For this we use OpenCV2’s COLOR_BGR2GRAY method and then use matplotlib’s cmap function

gray_puppies = cv2.cvtColor(puppies, cv2.COLOR_BGR2GRAY)

plt.figure(figsize = (9,7))
plt.imshow(gray_puppies, cmap='gray')

2. OpenCV2 implementation

Displaying images with CV2’s built-in “imshow” method

Python script implementation is below. Let’s rename our filename to basics.py

Output displays both color and black and white images in a 600X600 windows.

Takeaway things from basics.py script:

• script has 2 functions
• function 1: outputs color and black and white images in the same window if closed one after the other
• function 2: outputs both color and black and white images in two different windows
• order of function calls in the script:
  • 1. same_output_window()
  • 1. different_output_windows()

import cv2

def same_output_window():
    color_puppies = cv2.imread('puppies.jpg')
    cv2.namedWindow("output", cv2.WINDOW_NORMAL)
    cv2.resizeWindow('output', (600,600))
    cv2.imshow('output', color_puppies)  
    cv2.waitKey(0)

    blackandwhite_puppies = cv2.cvtColor(color_puppies, cv2.COLOR_BGR2GRAY)
    cv2.namedWindow("output", cv2.WINDOW_NORMAL)
    cv2.resizeWindow('output', (600,600))
    cv2.imshow('output', blackandwhite_puppies)  
    cv2.waitKey(0)

    cv2.destroyAllWindows()
    
def different_output_windows():
    color_puppies = cv2.imread('puppies.jpg')
    cv2.namedWindow("color", cv2.WINDOW_NORMAL)
    cv2.resizeWindow('color', (600,600))
    cv2.imshow('color', color_puppies)  

    blackandwhite_puppies = cv2.cvtColor(color_puppies, cv2.COLOR_BGR2GRAY)
    cv2.namedWindow("blackandwhite", cv2.WINDOW_NORMAL)
    cv2.resizeWindow('blackandwhite', (600,600))
    cv2.imshow('blackandwhite', blackandwhite_puppies)  
    cv2.waitKey(0)

    cv2.destroyAllWindows()
    
same_output_window()
different_output_windows()

Running python script from jupyter notebook

To run python script in jupyter, use the jupyter’s inline magic function %run to run

%run basics.py

Output of the script

Output of same_output_window() function that outputs loaded image with a same window name.

Notice highlighted markers: both windows are named “output”

Output of same_output_window() function that outputs loaded image with a different window name for both color and black and white image.

Notice highlighted markers: “color” window, “blackandwhite” window