Test easyocr

bad_ocr.py

@@ -0,0 +1,60 @@
+#!/usr/bin/env python3
+"""Extract text from images."""
+
+import os
+import easyocr
+import argparse
+import tqdm
+from toolz import compose, curry
+from toolz.curried import map, filter
+
+
+IMG_EXT = (
+    '.png',
+    '.jpeg',
+    '.jpg',
+    '.apng',
+    '.webp',
+    '.avif'
+)
+
+def list_images(folder):
+    _, _, files = next(os.walk(folder))
+    return compose(
+        map(lambda f: os.path.join(folder, f)), 
+        filter(lambda f: os.path.splitext(f)[-1].lower() in IMG_EXT)
+    )(files)
+
+@curry
+def ocr(reader, img_filename):
+    result = reader.readtext(img_filename, detail=0)
+    basename, _ = os.path.splitext(img_filename)
+    txt_filename = basename + '.txt'
+    with open(txt_filename, 'w') as f:
+        f.write(' '.join(result) + '\n')
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument(
+        'dir',
+        help="input dir"
+    )
+    parser.add_argument(
+        '-l', '--lang',
+        default='no,en',
+        help="comma-separated list of language codes"
+    )
+    args = parser.parse_args()
+
+    print("Loading model ...")
+    ocr_reader = easyocr.Reader(args.lang.split(','))
+    print("Run") 
+    compose(
+        list,
+        tqdm.tqdm,
+        map(ocr(ocr_reader)),
+        list_images
+    )(args.dir)
+    print("Done.")
+

requirements.pip

@@ -0,0 +1,24 @@
+cycler==0.10.0
+decorator==4.4.2
+easyocr==1.3.2
+imageio==2.9.0
+kiwisolver==1.3.1
+matplotlib==3.4.2
+networkx==2.5.1
+numpy==1.21.1
+opencv-python==4.5.3.56
+Pillow==8.3.1
+pyparsing==2.4.7
+python-bidi==0.4.2
+python-dateutil==2.8.2
+PyWavelets==1.1.1
+PyYAML==5.4.1
+scikit-image==0.18.2
+scipy==1.7.0
+six==1.16.0
+tifffile==2021.7.2
+toolz==0.11.1
+torch==1.9.0
+torchvision==0.10.0
+tqdm==4.61.2
+typing-extensions==3.10.0.0