easy_ocr.py

# %%
"""
## Importing Necessary Libraries 
"""

# %%
# !pip install git+git://github.com/jaidedai/easyocr.git
# !pip install easyocr

import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import cv2
from PIL import Image
import easyocr
import numpy as np
import math
import re
from scipy.spatial import distance

# %%
# import pkg_resources
# pkg_resources.get_distribution('scikit-learn').version

# %%
reader = easyocr.Reader(['en'])

# %%
"""
## Reading a sample Image and applying OCR
"""

# %%
# 9
image_path = "data/cnic_5.jpg"
im = Image.open(image_path)
image = cv2.imread(image_path)
result = reader.readtext(image_path)

# %%
dpi = im.info.get('dpi')
print(f"The DPI of Image is : {dpi}")

# %%
shape = image.shape[:-1]
print(f"The Shape of Image is :{shape}")

# %%
plt.imshow(image)

# %%
"""
## Showing the marked bounding boxes 
"""

# %%
bbs = []
values = []
# image = cv2.imread("data/cnic_2.jpg")
for (bbox, text, prob) in result:
    # display the OCR'd text and associated probability
#     print("[INFO] {:.4f}: {}".format(prob, text))
    values.append(text)
    
    # unpack the bounding box
    (tl, tr, br, bl) = bbox
    bbs.append(bbox)
    
    tl = (int(tl[0]), int(tl[1]))
    tr = (int(tr[0]), int(tr[1]))
    br = (int(br[0]), int(br[1]))
    bl = (int(bl[0]), int(bl[1]))
    
    # cleanup the text and draw the box surrounding the text along
    # with the OCR'd text itself
    cv2.rectangle(image, tl, br, (0, 255, 0), 2)
    cv2.putText(image, text, (tl[0], tl[1] - 10),
        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 0, 0), 2)
# show the output image
cv2.imshow("Image", image)
cv2.waitKey(0)

# %%
"""
## Normalizing the bounding Boxes 
"""

# %%
def normalize(img,result):
    w,h = img.shape[:-1]
    normalize_bbx = []
    detected_labels = []
    for (bbox, text, prob) in result:
        (tl, tr, br, bl) = bbox
        tl[0],tl[1] = round(tl[0] / h,3),round(tl[1] / w,3)
        tr[0],tr[1] = round(tr[0] / h,3),round(tr[1] / w,3)
        br[0],br[1] = round(br[0] / h,3),round(br[1] / w,3)
        bl[0],bl[1] = round(bl[0] / h,3),round(bl[1] / w,3)
        normalize_bbx.append([tl,tr,br,bl])
        detected_labels.append(text)
    return normalize_bbx,detected_labels

# %%
# CNIC 8
norm_boxes,labels = normalize(image,result)

# %%
normalize_output = list(zip(norm_boxes,labels)) 

# %%
"""
## Measuring Distance 
"""

# %%
def calculate_distance(key,bbx):
    euc_sum = 0
    for val1,val2 in zip(key,bbx):
        euc_sum = euc_sum + distance.euclidean(val1,val2)
        return euc_sum

# %%
"""
## Defining our Static Card Template Boxes  
"""

# %%
# name_key = [[0.272, 0.233], [0.323, 0.233], [0.323, 0.27], [0.272, 0.27]]
name_value = [[0.283, 0.271], [0.415, 0.271], [0.415, 0.325], [0.283, 0.325]]
# father_key = [[0.285, 0.42], [0.388, 0.42], [0.388, 0.457], [0.285, 0.457]]
father_value = [[0.29, 0.456], [0.494, 0.456], [0.494, 0.514], [0.29, 0.514]]
# dob_key = [[0.519, 0.713], [0.631, 0.713], [0.631, 0.756], [0.519, 0.756]]
dob_value = [[0.529, 0.751], [0.648, 0.751], [0.648, 0.803], [0.529, 0.803]]
# doi_key = [[0.274, 0.821], [0.384, 0.821], [0.384, 0.858], [0.274, 0.858]]
doi_value = [[0.285, 0.857], [0.404, 0.857], [0.404, 0.908], [0.285, 0.908]]
# doe_key = [[0.519, 0.821], [0.647, 0.821], [0.647, 0.866], [0.519, 0.866]]
doe_value = [[0.531, 0.859], [0.65, 0.859], [0.65, 0.911], [0.531, 0.911]]

# %%
"""
## Distances Comparison 
"""

# %%
def get_value(key,normalize_output):
    distances = {}
    for bbx,text in normalize_output:
        distances[text] = calculate_distance(key,bbx)
    return distances   

# %%
dict_data = {}
output_dict = {}
output_dict['Name'] = name_value
output_dict['Father Name']  = father_value
output_dict['Date of Birth'] = dob_value
output_dict['Date of Issue'] = doi_value
output_dict['Date of Expiry'] = doe_value

# %%
for key,value in output_dict.items():
    output_dict = get_value(value,normalize_output)
    answer = list(min(output_dict.items(), key=lambda x: x[1]))[0]
    dict_data[key] = answer 

# %%
"""
## Output Dictionary 
"""

# %%
dict_data

# %%
accuracy_df = pd.read_excel("data/accuracy.xlsx")

# %%
accuracy_df

# %%
x = accuracy_df["Accuracy (%)"]
y = accuracy_df["Shape (H,W)"]
z = accuracy_df["DPI"]

# %%
plt.figure(figsize=(30, 6))

plt.subplot(131)
plt.xlabel("Image Shape")
plt.ylabel("Accuracy (%)")
plt.plot(y,sorted(x))

# .lineplot(y,x)
plt.title("Image Shape Vs Accuracy")
plt.subplot(132)
sns.lineplot(z,x)
# plt.plot(z,x)
plt.title("Image DPI Vs Accuracy")
plt.show()

# %%


# %%
# !pip install streamlit