tensorflow实战之验证码识别

1.代码:

from captcha.image import ImageCaptcha
import matplotlib.pyplot as plt
from PIL import Image
import random
import tensorflow as tf
import numpy as np

numbers = ['0','1','2','3','4','5','6','7','8','9']
lower_case_letter = ['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z']
upper_case_letter = ['A','B','C','D','E','F','G','H','I','J','K','L','M','N','O','P','Q','R','S','T','U','V','W','X','Y','Z']

MAX_CAPTCHA=4
char_set=numbers

#随机生成验证码文字
def random_captcha_test(char_set=char_set,captcha_size=MAX_CAPTCHA):
    text=""
    for i in range(captcha_size):
        c=random.choice(char_set)
        text+=c
    return text

def gen_captcha():
    image=ImageCaptcha()
    captcha_text=random_captcha_test(numbers)
    captcha=image.generate(captcha_text)
    captcha_image=Image.open(captcha)
    captcha_image=np.array(captcha_image)
    return captcha_text,captcha_image

#RGB图转灰度图
def convert2gray(img):
    #print(img.shape)
    if len(img.shape)>2:
        gray=np.mean(img,2)
        # 上面的转法较快，正规转法如下
        # r, g, b = img[:,:,0], img[:,:,1], img[:,:,2]
        # gray = 0.2989 * r + 0.5870 * g + 0.1140 * b
        #Image.fromarray(gray,'L').save(text+'.jpg')
        return gray
    else:
        return img

#文本转向量
def text2vector(text):
    if len(text)>MAX_CAPTCHA:
        raise Exception("验证码超出最大长度")
    vector=np.zeros(len(char_set)*MAX_CAPTCHA)

    """
    ascll值
    0-9(0-9): 48-57
    A-Z(10-35): 65-90
    a-z(36-61): 97-122
    """
    def char2pos(c):
        if c =='_':
            k = 62
            return k
        k = ord(c)-48
        if k > 9:#字母
            k = ord(c) - 55
            if k > 35:#小写字母
                k = ord(c) - 61
                if k > 61:
                    raise ValueError('No Map')
        return k

    for i,c in enumerate(text):
        idx=i*MAX_CAPTCHA+char2pos(c)
        vector[idx]=1
    return vector

#向量转回文本
def vec2text(vec):
    char_pos = vec.nonzero()[0]
    text=[]
    for i, c in enumerate(char_pos):
        char_at_pos = i #c/63
        char_idx = c % len(char_set)
        if char_idx < 10:
            char_code = char_idx + ord('0')
        elif char_idx <36:
            char_code = char_idx - 10 + ord('A')
        elif char_idx < 62:
            char_code = char_idx-  36 + ord('a')
        elif char_idx == 62:
            char_code = ord('_')
        else:
            raise ValueError('error')
        text.append(chr(char_code))
    return "".join(text)


# 生成一个训练batch
def get_next_batch(batch_size=128):
    batch_x = np.zeros([batch_size, IMAGE_HEIGHT * IMAGE_WIDTH])
    batch_y = np.zeros([batch_size, MAX_CAPTCHA * len(char_set)])

    # 有时生成图像大小不是(60, 160, 3)
    def wrap_gen_captcha_text_and_image():
        while True:
            text, image = gen_captcha()
            if image.shape == (60, 160, 3):
                return text, image

    for i in range(batch_size):
        text, image = wrap_gen_captcha_text_and_image()
        image = convert2gray(image)

        batch_x[i, :] = image.flatten() / 255  # (image.flatten()-128)/128  mean为0
        batch_y[i, :] = text2vector(text)
    return batch_x, batch_y

#定义CNN
def crack_captcha_cnn(w_alpha=0.01, b_alpha=0.1):
    x=tf.reshape(X,(-1,IMAGE_HEIGHT,IMAGE_WIDTH,1))

    """
    X:60*160*1
    
    layer1:3*3*1 -> 32
    pool1:2*2 -> 30*80
    layer2:3*3*32 -> 64
    pool2:2*2 -> 15*40
    layer3:3*3*64 -> 64
    pool3:2*2 -> 8*20(SAME)
    full:8*20
    
    Y:40*1
    """
    w_c1 = tf.Variable(w_alpha * tf.random_normal([3, 3, 1, 32]))
    b_c1 = tf.Variable(b_alpha * tf.random_normal([32]))
    conv1 = tf.nn.relu(tf.nn.bias_add(tf.nn.conv2d(x, w_c1, strides=[1, 1, 1, 1], padding='SAME'), b_c1))
    conv1 = tf.nn.max_pool(conv1, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')
    conv1 = tf.nn.dropout(conv1, keep_prob)

    w_c2 = tf.Variable(w_alpha * tf.random_normal([3, 3, 32, 64]))
    b_c2 = tf.Variable(b_alpha * tf.random_normal([64]))
    conv2 = tf.nn.relu(tf.nn.bias_add(tf.nn.conv2d(conv1, w_c2, strides=[1, 1, 1, 1], padding='SAME'), b_c2))
    conv2 = tf.nn.max_pool(conv2, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')
    conv2 = tf.nn.dropout(conv2, keep_prob)

    w_c3 = tf.Variable(w_alpha * tf.random_normal([3, 3, 64, 64]))
    b_c3 = tf.Variable(b_alpha * tf.random_normal([64]))
    conv3 = tf.nn.relu(tf.nn.bias_add(tf.nn.conv2d(conv2, w_c3, strides=[1, 1, 1, 1], padding='SAME'), b_c3))
    conv3 = tf.nn.max_pool(conv3, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')
    conv3 = tf.nn.dropout(conv3, keep_prob)

    # Fully connected layer
    w_d = tf.Variable(w_alpha * tf.random_normal([8 * 20 * 64, 1024]))
    b_d = tf.Variable(b_alpha * tf.random_normal([1024]))
    dense = tf.reshape(conv3, [-1, w_d.get_shape().as_list()[0]])
    dense = tf.nn.relu(tf.add(tf.matmul(dense, w_d), b_d))
    dense = tf.nn.dropout(dense, keep_prob)

    w_out = tf.Variable(w_alpha * tf.random_normal([1024, MAX_CAPTCHA * len(char_set)]))
    b_out = tf.Variable(b_alpha * tf.random_normal([MAX_CAPTCHA * len(char_set)]))
    out = tf.add(tf.matmul(dense, w_out), b_out)
    # out = tf.nn.softmax(out)
    return out


def crack_captcha(captcha_image):
    output = crack_captcha_cnn()
    saver = tf.train.Saver()
    with tf.Session() as sess:
        saver.restore(sess, "./model/crack_capcha.model-810")

        predict = tf.argmax(tf.reshape(output, [-1, MAX_CAPTCHA , len(char_set)]), 2)
        text_list,out = sess.run([predict,output], feed_dict={X: [captcha_image], keep_prob: 1})
        print(out)
        text = text_list[0].tolist()
        return text

    # 训练
def train_crack_captcha_cnn():
    out=crack_captcha_cnn()

    loss=tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=out,labels=Y))
    optimizer=tf.train.AdamOptimizer().minimize(loss)

    #1*40 -> 4*10
    predict = tf.reshape(out, [-1, MAX_CAPTCHA, len(char_set)])
    max_idx_p = tf.argmax(predict, 2)
    max_idx_l = tf.argmax(tf.reshape(Y, [-1, MAX_CAPTCHA, len(char_set)]), 2)
    correct_pred = tf.equal(max_idx_p, max_idx_l)
    accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))

    saver = tf.train.Saver()
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())

        step=0
        while True:
            batch_x, batch_y = get_next_batch(64)
            _,loss_=sess.run([optimizer, loss],feed_dict={X:batch_x,Y:batch_y,keep_prob:0.75})
            #print(step, loss_)
            # 每100 step计算一次准确率
            if step % 10 == 0:
                batch_x_test, batch_y_test = get_next_batch(100)
                acc = sess.run(accuracy, feed_dict={X: batch_x_test, Y: batch_y_test, keep_prob: 1.})
                print(step, acc)
                # 如果准确率大于50%,保存模型,完成训练
                if acc > 0.80:
                    saver.save(sess, "./model/crack_capcha.model", global_step=step)
                    break
            step += 1


"""
X:60*160*1

Y:4*10

"""
if __name__ == '__main__':

    train=0

    if train==1:
        IMAGE_WIDTH = 160
        IMAGE_HEIGHT = 60

        X = tf.placeholder(tf.float32, (None, IMAGE_WIDTH * IMAGE_HEIGHT))
        Y = tf.placeholder(tf.float32, (None, MAX_CAPTCHA * len(char_set)))
        keep_prob = tf.placeholder(tf.float32)  # dropout

        train_crack_captcha_cnn()
    else:
        IMAGE_WIDTH = 160
        IMAGE_HEIGHT = 60
        text, image = gen_captcha()
        image=convert2gray(image)
        image = image.flatten() / 255

        X = tf.placeholder(tf.float32, (None, IMAGE_WIDTH * IMAGE_HEIGHT))
        Y = tf.placeholder(tf.float32, (None,MAX_CAPTCHA * len(char_set)))
        keep_prob = tf.placeholder(tf.float32)  # dropout

        predict_text = crack_captcha(image)
        print("正确: {}  预测: {}".format(text, predict_text))