Python NumPy (Page: 3)

Go to Page:

NumPy image manipulations

NumPy array can work with standard images. Let’s consider some examples. First of all, we need to have few libraries available:

numpy - Munpy array library itself
matplotlib.pyplot - is a state-based interface to matplotlib. It provides a MATLAB-like way of plotting.
scipy - The SciPy library is one of the core packages that make up the SciPy stack. It provides many user-friendly and efficient numerical routines, such as routines for numerical integration, interpolation, optimization, linear algebra, and statistics.

And we also have a picture of lovely fish, which will be our experimental picture!

Fish picture for NumPy demo
This is a picture of Sparisoma aurofrenatum or rainbow parrot fish from Wikipedia
Original image: 750 x 335

Read image with NumPy

First of all we will read image into NumPy array with ndimage function, check parameters of this image and display it with Python matplotlib.pyplot library


import numpy as np
import matplotlib.pyplot as plt
from scipy import ndimage

img_array = ndimage.imread('22_fish.jpg') # read image into NumPy array    
print(img_array.shape)                    # (335, 750, 3) y=335, x=750 
                                          # 3 layers for R, G and B colours
                                          # B&W image have only one colour layer
print(type(img_array))                    # 
plt.imshow(img_array)                     # convert into displayable format 
plt.show()                                # display image

Fish picture displayed with matplotlib.pyplot library
Fish picture displayed with matplotlib.pyplot library in our script
Original image: 634 x 385

In all following scripts, we are assuming, that we load all necessary libraries and already read image into ‘img_array’

Image slicing

We can slice our array and every slice will be a part of our image. Just with standard slicing tools


sliced = img_array[200:, 350:, :]
plt.imshow(sliced)
plt.show()

Image slicing
Display slice [200:, 350:, :], from 200 to 335 by Y, from 350 to 750 by X and all colours
Original image: 584 x 254

Image mirroring

Using these slices we can mirroring image along X or Y axes, or both


x_mirror = img_array[::-1, :, :]
plt.imshow(x_mirror)
plt.show()

Image mirroring
Image mirroring over X axis by slice with negative step
Original image: 583 x 280

Simple Image compression

Using step for slicing bigger than 1 we can make compression along this axis by removing some parts of the image. This compression will loose some data, so it will be not very accurate. For more accurate compression we need to recalculate average colours as well.


x_compress = img_array[:, ::3, :]
plt.imshow(x_compress)
plt.show()

Image compression
Simple 3 fold image compression along X axis but taking [:, ::3, :] slice

Array splitting

It is also possible to split array into few other arrays with ‘split’ function. Axis- show the direction of splitting. Value 0 can be ommited.


y = np.split(img_array, 5, axis=0)
plt.imshow(y[3])
plt.show()

1/5 of image
Displaying slice of image over Y axis.
Original image: 548 x 113

Array concatenation

Splited as an array image can be concatenated back. Furthermore – it can be concatenated in different order which will result in changing slices.

In this example we will split colour axis and then concatenate only one colour – this will give us Back&White image for only one colour from palette. Let’s try to make B&W over Green colour of this fish.


z = np.split(img_array, 3, axis=2)
plt.imshow(np.concatenate((z[1],z[1],z[1]), axis=2))
plt.show()

Monochrome image
Red and Blue colour layers are removed and Green turns into Black and White colour
Original image: 562 x 282

Go to Page: 1; 2; 3; 4; 5; 6;

Published: 2021-10-04 11:48:19
Updated: 2021-11-14 08:41:55

Last 10 artitles

2023-12-25 01:48:38 - In-place and out-place Numpy functions
2023-12-25 00:58:08 - First order ordinary differential equations
2023-12-25 00:22:15 - How to increase swap space in Ubuntu
2023-12-24 04:03:12 - How to accelerate math calculations with Numba and Cython
2023-12-19 00:38:20 - 3 ways of function caching in Python
2023-12-19 00:16:55 - Accelerating Python
2023-12-16 12:26:55 - BioPython
2023-12-02 11:05:34 - Reading FITS files with astropy
2023-12-02 05:23:42 - Astropy
2023-12-01 00:33:16 - Circle length calculation with the Monte-Carlo method

9 popular artitles

Python Panda (3446)
Python NumPy (2983)
Python DICTionary HowTo (1594)
Python LIST HowTo (1475)
Python (1422)
Python Generators (1281)
Python: How to make absolute links in BeautifulSoup (1091)
Programming language (968)
NumPy: What the difference between mgrid and meshgrid (950)