Most efficient method for reading/writing image files as numpy array

Question

I am wondering what the most efficient methods for reading/writing image and PDF files as numpy arrays for processing.

So far I have seen scipy.ndimage.imread and using PIL and numpy, which yeild the following results:

import os
import glob
from scipy.ndimage import imread
from PIL import Image
import numpy as np
import timeit
iters = 2
def scipy_fun():
    for x in glob.glob("*.jpg"):
        px = imread(x)
def PIL_fun():
    for x in glob.glob("*.jpg"):
        with Image.open(x) as im:
            px = np.array(im)

print(timeit.Timer(scipy_fun).timeit(number=iters))
print(timeit.Timer(PIL_fun).timeit(number=iters))

running the script shows similar results with marginally better from scipy:

2.8794324089019234
3.0174482765699095

Are there any faster ways to do this?

Thanks! imageio appears to be a very good library. The output was scipy: ~4 secs imageio: ~3 secs PIL: ~4 secs — chase
– chase, Commented Nov 3, 2016 at 15:03

Ali · Accepted Answer · 2018-12-10 07:12:08Z

1

First, do this

pip install pdf2image

Then,

import numpy as np
from pdf2image import convert_from_path as read
import PIL
import cv2
#pdf in the form of numpy array to play around with in OpenCV or PIL
img = np.asarray(read('path to the pdf file')[0])#first page of pdf

edited Dec 10, 2018 at 7:12

Ali

2,7643 gold badges37 silver badges62 bronze badges

answered Dec 6, 2018 at 12:01

Santhosh Dhaipule Chandrakanth

1,1921 gold badge14 silver badges27 bronze badges

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

1 Comment

Muneeb Ahmad Khurram Over a year ago

This approach is hell a lot smoother and more efficient than first saving the img and then using opencv to read it, so well done Ali and Santhosh!

Collectives™ on Stack Overflow

Most efficient method for reading/writing image files as numpy array

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related