How to Convert an image to NumPy array and saveit to CSV file using Python?

Let’s see how to Convert an image to NumPy array and then save that array into CSV file in Python? First, we will learn about how to convert an image to a numpy ndarray. There are many methods to convert an image to ndarray, few of them are:

Method 1: Using PIL and NumPy library.

We will use PIL.Image.open() and numpy.asarray().

Example:

Python3



filter_none

edit
close

play_arrow

link
brightness_4
code

# import required libraries
from PIL import Image
import numpy as gfg
  
# read an image
img = Image.open('geeksforgeeks.jpg')
  
# convert image object into array
imageToMatrice = gfg.asarray(img)
  
# printing shape of image
print(imageToMatrice.shape)

chevron_right


Output:

(251, 335, 3)

Method 2: Using Matplotib library.

We will use matplotlib.image.imread() method.

Example:

Python3

filter_none

edit
close

play_arrow

link
brightness_4
code

# import library
from matplotlib.image import imread
  
# read an image
imageToMatrice  = imread('geeksforgeeks.jpg')
  
# show shape of the image
print(imageToMatrice.shape)

chevron_right


Output:

(251, 335, 3)

Now, imageToMatrice variable contains the ndarray which is obtained after the conversion from the given image. 

The dimension of the matrice obtained is decided by how many channels are present in the image:



  • For a black and white or gray scale image: There is only one channel present, thus, the shape of the matrices would be (n, n) where n represents the dimension of the images (pixels), and values inside the matrix range from 0 to 255.
  • For color or RGB image: It will render a tensor of 3 channels, thus the shape of the matrices would be (n, n,3). Each channel is an (n, n) matrix where each entry represents respectively the level of Red, Green, or Blue at the actual location inside the image.

We will use two methods to do the same, first method using numpy library and the second method using pandas library:

Note: we can save only 1D or 2D matrix in a file, therefore, there would be no issue in the gray scale or black and white image as it is a 2D matrix, but we need to make sure that this works for a colored or RGB image, which is a 3D matrix.

Method 1: Using NumPy library.

We will use numpy.savetxt() and numpy.loadtxt().

Example:

Python3

filter_none

edit
close

play_arrow

link
brightness_4
code

# import required libraries
import numpy as gfg
import matplotlib.image as img
  
# read an image
imageMat = img.imread('gfg.jpg')
print("Image shape:", imageMat.shape)
  
# if image is colored (RGB)
if(imageMat.shape[2] == 3):
    
  # reshape it from 3D matrice to 2D matrice
  imageMat_reshape = imageMat.reshape(imageMat.shape[0],
                                      -1)
  print("Reshaping to 2D array:",
        imageMat_reshape.shape)
  
# if image is grayscale
else:
  # remain as it is
  imageMat_reshape = imageMat
      
# saving matrice to .csv file
gfg.savetxt('geek.csv',
            imageMat_reshape)
  
# retrieving matrice from the .csv file
loaded_2D_mat = gfg.loadtxt('geek.csv')
  
# reshaping it to 3D matrice
loaded_mat = loaded_2D_mat.reshape(loaded_2D_mat.shape[0], 
                                   loaded_2D_mat.shape[1] // imageMat.shape[2],
                                   imageMat.shape[2])
  
print("Image shape of loaded Image:"
      loaded_mat.shape)
  
# check if both matrice have same shape or not
if((imageMat == loaded_mat).all()):
    
  print("\n\nYes",
        "The loaded matrice from CSV file is same as original image matrice")

chevron_right


Output:

Image shape: (251, 335, 3)
Reshaping to 2D array: (251, 1005)
Image shape of loaded Image: (251, 335, 3)

Yes The loaded matrice from CSV file is same as original image matrice

Method 2: Using Pandas library.

We will use pandas.Dataframe() and pandas.Dataframe() .to_csv() method.

Python3

filter_none

edit
close

play_arrow

link
brightness_4
code

# import required libraries
import numpy as gfg
import matplotlib.image as img
import pandas as pd
  
# read an image
imageMat = img.imread('gfg.jpg')
print("Image shape:"
      imageMat.shape)
  
# if image is colored (RGB)
if(imageMat.shape[2] == 3):
    
  # reshape it from 3D matrice to 2D matrice
  imageMat_reshape = imageMat.reshape(imageMat.shape[0],
                                      -1)
  print("Reshaping to 2D array:"
        imageMat_reshape.shape)
  
# if image is grayscale
else:
  # remain as it is
  imageMat_reshape = imageMat
      
# converting it to dataframe.
mat_df = pd.DataFrame(imageMat_reshape)
  
# exporting dataframe to CSV file.
mat_df.to_csv('gfgfile.csv',
              header = None,
              index = None)
  
# retrieving dataframe from CSV file
loaded_df = pd.read_csv('gfgfile.csv'
                        sep = ',',
                        header = None)
# getting matrice values.
loaded_2D_mat = loaded_df.values
  
# reshaping it to 3D matrice
loaded_mat = loaded_2D_mat.reshape(loaded_2D_mat.shape[0],
                                   loaded_2D_mat.shape[1] // imageMat.shape[2],
                                   imageMat.shape[2])
  
print("Image shape of loaded Image :",
      loaded_mat.shape)
  
# check if both matrice have same shape or not
if((imageMat == loaded_mat).all()):
  print("\n\nYes",
        "The loaded matrice from CSV file is same as original image matrice")

chevron_right


Output:

Image shape: (251, 335, 3)
Reshaping to 2D array: (251, 1005)
Image shape of loaded Image : (251, 335, 3)

Yes The loaded matrice from CSV file is same as original image matrice




My Personal Notes arrow_drop_up

Check out this Author's contributed articles.

If you like GeeksforGeeks and would like to contribute, you can also write an article using contribute.geeksforgeeks.org or mail your article to contribute@geeksforgeeks.org. See your article appearing on the GeeksforGeeks main page and help other Geeks.

Please Improve this article if you find anything incorrect by clicking on the "Improve Article" button below.