[SOLVED] Why is cv2.imread() not storing the name of my image files?

Why is cv2.imread() not storing the name of my image files?

I'm trying to write a script which will take all of the images in a given directory and format the size.

I have been able to import the files as a list using OS and split them using the loop. I've printed each file name and index successfully, but when I try to get the dimensional values using cv2.imread(), it returns 'None', making it impossible for me to get the shape, and throws me an AttributeError.

I have already tried uninstalling and reinstalling opencv-python. As suggested here.

import os
import cv2


def imageResize():
    dirlist = os.listdir('images')

    for c, file in enumerate(dirlist):
        print(c, file)
        img = cv2.imread(file)
        height, width = img.shape[0, 2]
        print(height, width)


if __name__ == '__main__':
    imageResize()

Expected Output:

0 image17_10.png
600px 400px

1 image15_9.png
500px 500px

...

Actual Output:

0 image17_10.png
height, width = img.shape[0,2]

AttributeError: 'NoneType' object has no attribute 'shape'

Solution

Welcome on stack overflow! It seems that you missed the path when opening the file, the line should be:

cv2.imread('images' + os.path.sep + file)
# or
cv2.imread(os.path.join('images', file))

or any other way of constructing the path to the file from the filename.