Captcha-Solver/CAPTCHAsolver.py at main · MingEn82/Captcha-Solver · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
import os
import xml.etree.cElementTree as ET
import csv
import numpy as np
import cv2
from imutils import paths
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelBinarizer
from keras.models import Sequential
from keras.layers.convolutional import Conv2D, MaxPooling2D
from keras.layers.core import Flatten, Dense, Dropout
from keras.callbacks import EarlyStopping

DATA_PATH = r"D:\Python projects\AutoTempTaking\CAPTCHAsolver\Data\Images"
XML_PATH = r"D:\Python projects\AutoTempTaking\CAPTCHAsolver\Data\xml"
OUTPUT_FOLDER = r"D:\Python projects\AutoTempTaking\CAPTCHAsolver\Data\extracted letters"


def xml_data_to_csv(xml_path, csv_path, savename):
    """
    Convert xml files to csv - deprecated
    """
    with open(os.path.join(csv_path, savename+'.csv'), 'w', newline='') as csvfile:
        fieldnames = ['filename', 'width', 'height', 'name', 'xmin', 'ymin', 'xmax', 'ymax']
        xmlwriter = csv.DictWriter(csvfile, fieldnames=fieldnames)
        xmlwriter.writeheader()
        for file in os.listdir(xml_path):
            if file.endswith('.xml'):
                tree = ET.parse(os.path.join(xml_path, file)).getroot()
                filename = tree.find('filename').text
                width = tree.find('size').find('width').text
                height = tree.find('size').find('height').text
                for char in tree.findall('object'):
                    name = char.find('name').text
                    xmin = char.find('bndbox').find('xmin').text
                    ymin = char.find('bndbox').find('ymin').text
                    xmax = char.find('bndbox').find('xmax').text
                    ymax = char.find('bndbox').find('ymax').text
                    xmlwriter.writerow({'filename': filename, 'width': width, 'height': height,
                                        'name': name, 'xmin': xmin, 'ymin': ymin, 'xmax': xmax, 'ymax': ymax})


def char_separator(DATA_PATH, XML_PATH, OUTPUT_FOLDER):
    """
    To separate the captcha images into individual letters for training and validating
    """
    counts = {}

    for img in os.listdir(DATA_PATH):
        image = cv2.imread(os.path.join(DATA_PATH, img))
        image = cv2.medianBlur(image, 5)

        # edge detection
        edges = cv2.Canny(image, 100, 200)
        edges = cv2.bitwise_not(edges)

        base_name = os.path.splitext(img)[0]
        tree = ET.parse(os.path.join(XML_PATH, base_name + '.xml')).getroot()
        for char in tree.findall('object'):
            name = char.find('name').text
            xmin = int(char.find('bndbox').find('xmin').text)
            ymin = int(char.find('bndbox').find('ymin').text)
            xmax = int(char.find('bndbox').find('xmax').text)
            ymax = int(char.find('bndbox').find('ymax').text)

            letter_image = edges[ymin:ymax, xmin:xmax]
            save_path = os.path.join(OUTPUT_FOLDER, name)

            # creating different output folder for storing different letters
            if not os.path.exists(save_path):
                os.makedirs(save_path)

            # write the letter image to a file
            count = counts.get(name, 1)
            p = os.path.join(save_path, "{}.png".format(str(count)))
            cv2.imwrite(p, letter_image)

            # increment the count
            counts[name] = count + 1


data = []
labels = []
for image in paths.list_images(OUTPUT_FOLDER):
    # pre-processing images - may not be necessary as I already did some
    img = cv2.imread(image)
    img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    # standardise images to 30x30
    img = cv2.resize(img, (30,30))
    # expand dimensions of image from 30x30 to 30x30x1 - Tensorflow requirement
    img = np.expand_dims(img, axis=2)

    label = image.split(os.path.sep)[-2]
    data.append(img)
    labels.append(label)

data = np.array(data, dtype='float')
labels = np.array(labels)

# normalise the points
data = data/255.0
(train_imgs, test_imgs, train_labels, test_labels) = train_test_split(data, labels, test_size=0.2, random_state=0)

# labels index are based on label file position
lb = LabelBinarizer().fit(train_labels)
train_labels = lb.transform(train_labels)
test_labels = lb.transform(test_labels)

model = Sequential()
model.add(Conv2D(20, (5, 5), padding="same", input_shape=(30, 30, 1), activation="relu"))
model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2)))
model.add(Conv2D(50, (5, 5), padding="same", activation="relu"))
model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2)))
model.add(Flatten())
model.add(Dense(128, activation="relu"))
model.add(Dropout(0.3))
model.add(Dense(28, activation="softmax"))

model.compile(loss="categorical_crossentropy", optimizer="adam", metrics=["accuracy"])

# using early stopping for avoiding overfitting
estop = EarlyStopping(patience=10, mode='min', min_delta=0.001, monitor='val_loss')

model.fit(train_imgs, train_labels, validation_data=(test_imgs, test_labels), batch_size=32, epochs=50, verbose=1, callbacks = [estop])

model.save("model.h5")

# loads model for future use
# model = keras.models.load_model("model.h5")