server_web.py

from flask import Flask, render_template, request, send_from_directory, send_file, redirect
from werkzeug.utils import secure_filename
from transform import four_point_transform
import cv2
import numpy as np
import imutils
import argparse
from skimage.filters import threshold_local
import math
from PIL import Image
import pytesseract

app = Flask(__name__)

def thresholding(img):
    dilated_img = cv2.dilate(img, np.ones((5,5), np.uint8))
    bg_img = cv2.medianBlur(dilated_img, 21)
    diff_img = 255 - cv2.absdiff(img, bg_img)
    norm_img = diff_img.copy()  # Needed for 3.x compatibility
    cv2.normalize(diff_img, norm_img, alpha=0, beta=255, norm_type=cv2.NORM_MINMAX, dtype=cv2.CV_8UC1)
    _, thr_img = cv2.threshold(norm_img, 230, 0, cv2.THRESH_TRUNC)
    cv2.normalize(thr_img, thr_img, alpha=0, beta=255, norm_type=cv2.NORM_MINMAX, dtype=cv2.CV_8UC1)
    return thr_img

def angle(x1,y1,x2,y2,x3,y3):
    if x2==x1:
        if x3==x2:
            ang=0
        else:
            m2 = (y3 - y2) / (x3 - x2)
            if m2==0:
                ang = 1.57
            else:
                ang = math.atan(1/m2)
    elif x2 == x3:
        if x1==x2:
            ang=0
        else:
            m1 = (y2 - y1) / (x2 - x1)
            if m1==0:
                ang = 1.57
            else:
                ang = math.atan(-1/m1)
    else:
        m1 = (y2-y1)/(x2-x1)
        m2 = (y3-y2)/(x3-x2)
        # print(m1,m2)
        if (m1*m2) != -1:
            m = (m1-m2)/(1+(m1*m2))
            ang = math.atan(m)
        else:
            ang = 1.57
    if ang<0:
        ang = ang+3.14

    return ang

def document_scanner(original_image,heights):
    global original
    original = original_image.copy()
    for height in heights:

        ratio = original_image.shape[0] /height
        original_image = imutils.resize(original_image, height=height)
        gray = cv2.cvtColor(original_image , cv2.COLOR_BGR2GRAY)
        gray = cv2.GaussianBlur(gray,(5,5),0)
        # gray = cv2.adaptiveThreshold(gray , 255 , cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,10,0)
        # _, gray = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
        gray = thresholding(original_image)
        # cv2.imshow('Gray2',gray)
        print('Height',height)
        threshold = [0,50,100]
        for i in threshold:
            edged = cv2.Canny(original_image,i,i+50)
            print('{}  {}'.format(i,i+50))
        # print(gray.shape)
        # print(img.shape)
        #     print(edged.shape)

        # cv2.imshow('Gray' , gray)
        # cv2.imshow('Edged' , edged)
        # cv2.waitKey(0)
        # cv2.destroyAllWindows()

            cnts = cv2.findContours(edged.copy() , cv2.RETR_EXTERNAL , cv2.CHAIN_APPROX_SIMPLE)
            cnts = imutils.grab_contours(cnts)
            cnts = sorted(cnts, key = cv2.contourArea, reverse = True)[:5]
            check =False
            print('True1')

            for c in cnts:
                peri = cv2.arcLength(c, True)
                approx = cv2.approxPolyDP(c, 0.02 * peri, True)

                if len(approx) == 4:
                    screenCnt = approx
                    check = True
                    break

            print(screenCnt)
            print('Sreencnt',screenCnt[0][0][0])
            print(screenCnt[1][0])
            print(screenCnt[1][0][0])
            a1 = angle(screenCnt[0][0][0],screenCnt[0][0][1],screenCnt[1][0][0],screenCnt[1][0][1],screenCnt[2][0][0],screenCnt[2][0][1])
            a2 = angle(screenCnt[1][0][0],screenCnt[1][0][1],screenCnt[2][0][0],screenCnt[2][0][1],screenCnt[3][0][0],screenCnt[3][0][1])
            a3 = angle(screenCnt[2][0][0],screenCnt[2][0][1],screenCnt[3][0][0],screenCnt[3][0][1],screenCnt[0][0][0],screenCnt[0][0][1])
            a4 = angle(screenCnt[3][0][0],screenCnt[3][0][1],screenCnt[0][0][0],screenCnt[0][0][1],screenCnt[1][0][0],screenCnt[1][0][1])

            if a1<1.918 and a1>1.222 and a2<1.918 and a2>1.222 and a3<1.918 and a3>1.222 and a4<1.918 and a4>1.222:
                if check:
                    print("STEP 2: Find contours of paper")
                    cv2.drawContours(original_image, [screenCnt], -1, (0, 255, 0), 2)
                    # cv2.imshow("Outline", original_image)
                    # cv2.waitKey(0)
                    # cv2.destroyAllWindows()
                    break
                else:
                    continue
        if check:
            break
        else:
            continue

    if a1<1.918 and a1>1.222 and a2<1.918 and a2>1.222 and a3<1.918 and a3>1.222 and a4<1.918 and a4>1.222:
        print('True2')
        warped = four_point_transform(original, screenCnt.reshape(4, 2) *ratio)
        warped = cv2.cvtColor(warped, cv2.COLOR_BGR2GRAY)
        # _,warped = cv2.threshold(warped , 0 , 255, cv2.THRESH_BINARY+cv2.THRESH_OTSU)
        # cv2.imshow('YOY',warped)
        T = threshold_local(warped, 11, offset=10, method="gaussian")
        warped = (warped > T).astype("uint8") * 255
    else:
        print('True3')
        warped = original
        warped = cv2.cvtColor(warped , cv2.COLOR_BGR2GRAY)
        T = threshold_local(warped ,15,offset=4,method='gaussian')
        warped = (warped>T).astype('uint8')*255
    return warped


@app.route('/')
@app.route('/upload_doc')
def upload():

    return render_template('upload_document.html')

@app.route('/scanned', methods=['GET', 'POST'])
def show():
    if request.method == 'POST':
        image = request.files['file']
        print(image)
        filename = secure_filename(image.filename)
        # image.save('C:/users/spars/desktop/flask_catdog/templates/'+filename)
        open('templates/' + 'uploaded.jpg', 'wb').write(image.read())
        img = cv2.imread('templates/uploaded.jpg')
        try:
            if img.shape[1] > 3000:
                warped = document_scanner(img, [1500, 1300])
            elif img.shape[1] <= 3000:
                warped = document_scanner(img, [500, 700, 1200])
        except:
            print("Image Not Captured Properly")
            warped = img
            warped = cv2.cvtColor(warped, cv2.COLOR_BGR2GRAY)
            T = threshold_local(warped, 15, offset=4, method='gaussian')
            warped = (warped > T).astype('uint8') * 255

        # print("STEP 3: Apply perspective transform")
        # cv2.imshow("Original", imutils.resize(original, height=650))
        # cv2.imshow("Scanned", imutils.resize(warped, height=650))
        filename = filename.split('.')[0]+'.jpg'
        cv2.imwrite('templates/'+filename, warped)
        response = {
            'img_name' : filename
        }
        text = pytesseract.image_to_string(Image.open('templates/'+filename))
        response['text'] = text
        return render_template('show_scanned_doc.html' , response = response)

@app.route('/upload/<filename>')
def send_file(filename):
    return send_from_directory('templates',filename)

if __name__ == '__main__':
    app.run( debug=True)