fixup.py

from PIL import Image
from glob import glob
import os

# pip install imagehash
import imagehash

images = {}

for f in glob("images/*"):
    i = Image.open(f)
    if i.size[0] < i.size[1]:
        print f
        i = i.rotate(90)
        i.save(f)

    # check for dupe-ish images. 5 is a sensitivity parameter; its value was
    # determined empirically.
    h = imagehash.average_hash(i, 5).hash.tostring()
    if h in images:
        print "{} is a dupe of {}? deleting.".format(f, images[h])
        os.unlink(f)
    else:
        images[h] = f