image_utils.py

"""
MIT License

Copyright (c) Research @ RedHunt Labs Pvt Ltd
Written by Owais Shaikh
Email: owais.shaikh@redhuntlabs.com | me@0x4f.in

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
"""

import cv2, text_utils, pytesseract, difflib
from PIL import Image

from skimage.transform import rotate
from deskew import determine_skew
import numpy

def scan_image_for_people(image):
    
    image = numpy.array(image) # converts the image to a compatible format
    cascade_values_file = 'face_cascade.xml'
    cascade_values = cv2.CascadeClassifier(cascade_values_file)
    faces = cascade_values.detectMultiScale (
        image,
        scaleFactor = 1.1,
        minNeighbors = 5,
        minSize = (30, 30),
        flags = cv2.CASCADE_SCALE_IMAGE
    )

    return len(faces) 

def scan_image_for_text(image):
    image = numpy.array(image) # converts the image to a compatible format

    # 0. Original 
    #print ("Reading text from unmodified image")
    try:
        image_text_unmodified = pytesseract.image_to_string(image, config = '--psm 12')
    except TypeError:
        print("Cannot open this file type.")
        return
    #cv2.imwrite("image0.png", image)

    # 1. Auto-rotation
    try: 
        #print ("Attempting to auto-rotate image and read text")
        try:
            degrees_to_rotate = pytesseract.image_to_osd(image)
        except: degrees_to_rotate = "Rotate: 180"

        for item in degrees_to_rotate.split("\n"):
            if "rotate".lower() in item.lower():
                degrees_to_rotate = int(item.replace(" ", "").split(":", 1)[1])
                if degrees_to_rotate == 180:
                    pass
                elif degrees_to_rotate == 270:
                    image = cv2.rotate(image, cv2.ROTATE_90_COUNTERCLOCKWISE)
                elif degrees_to_rotate == 360:
                    image = cv2.rotate(image, cv2.ROTATE_180)

        image_text_auto_rotate = pytesseract.image_to_string(image, config = '--psm 12')
        #cv2.imwrite("image1.png", image)
    except: 
        print ("Couldn't auto-rotate image")
        image_text_auto_rotate = ""

    # 2. Grayscaled
    try: 
        #print ("Reading text from grayscaled image")
        image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
        image_text_grayscaled = pytesseract.image_to_string(image, config = '--psm 12')
        #cv2.imwrite("image2.png", image)
    except: 
        print ("Couldn't grayscale image")
        image_text_grayscaled = ""

    # 3. Monochromed
    try: 
        #print ("Reading text from monochromed image")
        image = cv2.threshold(image, 100, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)[1]
        image_text_monochromed = pytesseract.image_to_string(image, config = '--psm 12')
        #cv2.imwrite("image3.png", image)
    except: 
        print ("Couldn't monochrome image")
        image_text_monochromed = ""

    # 4. Mean threshold
    try:
        #print ("Reading text from mean threshold image")
        image = cv2.adaptiveThreshold(image, 255, cv2.ADAPTIVE_THRESH_MEAN_C, cv2.THRESH_BINARY, 11, 2)
        image_text_mean_threshold = pytesseract.image_to_string(image, config = '--psm 12')
        #cv2.imwrite("image4.png", image)
    except: 
        print ("Couldn't mean threshold image")
        image_text_mean_threshold = ""

    # 5. Gaussian threshold
    try:
        #print ("Reading text from gaussian threshold image")
        image = cv2.adaptiveThreshold(image, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY, 11, 2)
        image_text_gaussian_threshold = pytesseract.image_to_string(image, config = '--psm 12')
        #cv2.imwrite("image5.png", image)
    except: 
        print ("Couldn't gaussian threshold image")
        image_text_gaussian_threshold = ""

    # 6. Deskew
    try:
        # 6a. Deskew one
        #print ("First attempt at de-skewing image and reading text")
        angle = determine_skew(image)
        rotated = rotate(image, angle, resize=True) * 255
        image = rotated.astype(numpy.uint8)
        image_text_deskewed_1 = pytesseract.image_to_string(image, config = '--psm 12')
        #cv2.imwrite("image6.png", image)

        # 6b. Deskew two
        #print ("Second attempt at de-skewing image and reading text")
        angle = determine_skew(image)
        rotated = rotate(image, angle, resize=True) * 255
        image = rotated.astype(numpy.uint8)
        image_text_deskewed_2 = pytesseract.image_to_string(image, config = '--psm 12')
        #cv2.imwrite("image7.png", image)

        # 6c. Deskew three
        #print ("Third attempt at de-skewing image and reading text")
        angle = determine_skew(image)
        rotated = rotate(image, angle, resize=True) * 255
        image = rotated.astype(numpy.uint8)
        image_text_deskewed_3 = pytesseract.image_to_string(image, config = '--psm 12')
        #cv2.imwrite("image8.png", image)
    except: 
        print ("Couldn't deskew image")
        image_text_deskewed_1 = ""
        image_text_deskewed_2 = ""
        image_text_deskewed_3 = ""

    # END OF IMAGE PROCESSING AND OCR

    # Tokenize all scanned strings by newlines and spaces
    unmodified_words = text_utils.string_tokenizer(image_text_unmodified)
    grayscaled = text_utils.string_tokenizer(image_text_grayscaled)
    auto_rotate = text_utils.string_tokenizer(image_text_auto_rotate)
    monochromed = text_utils.string_tokenizer(image_text_monochromed)
    mean_threshold = text_utils.string_tokenizer(image_text_mean_threshold)
    gaussian_threshold = text_utils.string_tokenizer(image_text_gaussian_threshold)
    deskewed_1 = text_utils.string_tokenizer(image_text_deskewed_1)
    deskewed_2 = text_utils.string_tokenizer(image_text_deskewed_2)
    deskewed_3 = text_utils.string_tokenizer(image_text_deskewed_3)

    original = image_text_unmodified + "\n" + image_text_auto_rotate + "\n" + image_text_grayscaled + "\n" + image_text_monochromed + "\n" + image_text_mean_threshold + "\n" + image_text_gaussian_threshold + "\n" + image_text_deskewed_1 + "\n" + image_text_deskewed_2 + "\n" +  image_text_deskewed_3

    intelligible = unmodified_words + grayscaled + auto_rotate + monochromed + mean_threshold + gaussian_threshold + deskewed_1 + deskewed_2 + deskewed_3

    return (original, intelligible)