1

I want to crop images with different sizes to get the same size to futher process them. I wrote the following code:

import glob
import cv2
import os
from matplotlib import pyplot as plt


inputFolder = "C:\\Users\\die5k\\Desktop\\hist\\Cropping\\input"
storeDirectory =  "C:\\Users\\die5k\\Desktop\\hist\\Cropping\\output"

path = glob.glob(inputFolder + "\\*.png")
cv_img = []
image_no = 1
for img in path:
    n = cv2.imread(img)
    cv_img.append(n)
    print(img)

    os.chdir(storeDirectory)
    cropped_img = n.crop(((w-100)//2, (h-100)//2, (w+100)//2, (h+100)//2))


    filename = "Figure_" + str(image_no) + ".png"

    plt.gcf().savefig(filename)
   

    print(image_no)
    image_no += 1

This outputs me the following error: AttributeError: 'numpy.ndarray' object has no attribute 'crop'

I am coding beginner and I dont know what I have to do.

Tapsi
  • 39
  • 1
  • 4

1 Answers1

2

It's because numpy doesn't have crop functionality. Try opening the image using PIL library and use the crop function as follows:

from PIL import Image    
n = Image.open(path)

And then proceed with the crop. Or Alternatively, you can crop it yourself without the function as follows:

cropped_img = n[((h-100)//2):((h-100)//2)+((h+100)//2), ((w-100)//2):((w-100)//2)+((w+100)//2)]
Raghul Raj
  • 1,428
  • 9
  • 24