combine 3d arrays into a 4d array in numpy

Question

I have 3 images of the same size, I read each one using matplotlib.image.imread which returns a 3D numpy array.

(Height, Width, ChannelsNumber) # ChannelsNumber = 3 for RGB

Now, i want to merge the 3 images into a 4D numpy array like this :

(Height, Width, ChannelsNumber, NumberOfImages)

Meaning, the 3rd dimension array at position (X,Y) will no longer contain RGB values for pixel (X,Y) but will contain 3 arrays one for each channel (), where each array will contain 3 R, G, r B values (for 3 images)

array at position X,Y,0 will contain all RED values for images 1,2,3 at pixel X,Y
array at position X,Y,1 will contain all GREEN values for images 1,2,3 at pixel X,Y
array at position X,Y,2 will contain all BLUE values for images 1,2,3 at pixel X,Y

For example :

(10,15,0) will be an array containing all the RED values for all the images for Pixel (10,15)

Here's what i have tried :

import numpy as np
import os
import matplotlib.image as mpimg

img1 = mpimg.imread('./image1.jpg')
img2 = mpimg.imread('./image2.jpg')
img3 = mpimg.imread('./image3.jpg')

x = np.dstack((img1, img2, img3))

this returns a 3D array, where the 3rd dimension is an array that contains the RGB values of the images in this format:

[R-img1, B-img1, G-img1, R-img2, B-img2, G-img2, R-img3, B-img3, G-img3]

where what i want is :

[ 
    [R-img1, R-img2, R-img3,],
    [R-img1, B-img2, -img3,],
    [B-img1, B-img2, B-img3,]
]

Thank you !

@Scott thank you .. put this as an answer so that i mark this as answerd .. — Mohamed Benkedadra
– Mohamed Benkedadra, Commented Nov 10, 2019 at 1:13
@Scott how do i reverse, do the opposite, for output to input ? — Mohamed Benkedadra
– Mohamed Benkedadra, Commented Nov 10, 2019 at 10:01
The first img is the 0th element along the last index. img1 = stacked_imgs[...,0] — Him
– Him, Commented Nov 10, 2019 at 13:05

Mohamed Benkedadra · Accepted Answer · 2019-11-10 01:25:40Z

3

the answer is the one Scott commented

x = np.stack((img1, img2, img3), axis = -1)

answered Nov 10, 2019 at 1:25

Mohamed Benkedadra

2,0945 gold badges24 silver badges51 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Sebin Sunny · Accepted Answer · 2019-11-10 01:17:19Z

2

import numpy as np
import os
import matplotlib.image as mpimg

img1 = mpimg.imread('1.png')
img2 = mpimg.imread('2.png')
img3 = mpimg.imread('1.png')

print(img1.shape)
print(img2.shape)
print(img3.shape)
x = np.vstack((img1, img2, img3))

Stack arrays in sequence vertically

answered Nov 10, 2019 at 1:17

Sebin Sunny

1,81311 silver badges9 bronze badges

Collectives™ on Stack Overflow

combine 3d arrays into a 4d array in numpy

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related