Reading and writing RGB files using OpenCV

February 10, 2022

New to python and OpenCV. I’m using OpenCV to read a jpg, I then write it to a file as hex data, I read the file back again and construct a numpy array and use imshow to show the image, but it looks different from the original image. Why did the image color get blue?

file1.py:

import cv2
import numpy as np

img = cv2.imread('D:\cat.jpg')

cv2.imshow('img', img)     #Shows the image
cv2.waitKey(3000)

with open('D:\catHex', 'wb') as f:
    f.write(img)
    f.close()

file2.py:

import cv2
import numpy as np

img = np.zeros((640,640,3), np.uint8)
row = 0
col = 0

with open('D:\catHex', 'rb') as f:

    while(row < 640):

        img[row,col,2] = ord(f.read(1))
        img[row,col,1] = ord(f.read(1))
        img[row,col,0] = ord(f.read(1))

        col = col + 1
        if(col == 640):
            col = 0
            row = row + 1

    cv2.imshow('img',img)
    cv2.waitKey(3000)

>Solution :

You are swapping the Red and Blue channels when reading the image in your second script, therefore, the image looks blue. The following should work as expected:

img[row,col,0] = ord(f.read(1))
img[row,col,1] = ord(f.read(1))
img[row,col,2] = ord(f.read(1))

Also you could write your while loop in the following more pythonesque way with for-loops:

for row in range(640):
    for col in range(640):
        img[row,col,0] = ord(f.read(1))
        img[row,col,1] = ord(f.read(1))
        img[row,col,2] = ord(f.read(1))

    cv2.imshow('img',img)
    cv2.waitKey(3000)