from gen_check_code import gen_captcha_text_and_image_new,gen_captcha_text_and_imagefrom gen_check_code import numberfrom test_check_code import get_test_captcha_text_and_imageimport numpy as npimport tensorflow as tftext, image = gen_captcha_text_and_image_new()print("驗(yàn)證碼圖像channel:", image.shape) # (60, 160, 3) # 圖像大小 IMAGE_HEIGHT = image.shape[0]IMAGE_WIDTH = image.shape[1]image_shape = image.shapeMAX_CAPTCHA = len(text)print("驗(yàn)證碼文本最長字符數(shù)", MAX_CAPTCHA) # 驗(yàn)證碼最長4字符; 我全部固定為4,可以不固定. 如果驗(yàn)證碼長度小于4，用'_'補(bǔ)齊# 把彩色圖像轉(zhuǎn)為灰度圖像（色彩對識(shí)別驗(yàn)證碼沒有什么用）# 度化是將三分量轉(zhuǎn)化成一樣數(shù)值的過程def convert2gray(img): if len(img.shape) > 2:  gray = np.mean(img, -1)  # 上面的轉(zhuǎn)法較快，正規(guī)轉(zhuǎn)法如下   # r, g, b = img[:,:,0], img[:,:,1], img[:,:,2]   # gray = 0.2989 * r + 0.5870 * g + 0.1140 * b  # int gray = (int) (0.3 * r + 0.59 * g + 0.11 * b);  return gray else:  return img""" cnn在圖像大小是2的倍數(shù)時(shí)性能最高, 如果你用的圖像大小不是2的倍數(shù)，可以在圖像邊緣補(bǔ)無用像素。 np.pad(image,((2,3),(2,2)), 'constant', constant_values=(255,)) # 在圖像上補(bǔ)2行，下補(bǔ)3行，左補(bǔ)2行，右補(bǔ)2行 """char_set = number # 如果驗(yàn)證碼長度小于4, '_'用來補(bǔ)齊CHAR_SET_LEN = len(char_set)# 文本轉(zhuǎn)向量def text2vec(text): text_len = len(text) if text_len > MAX_CAPTCHA:  raise ValueError('驗(yàn)證碼最長4個(gè)字符') vector = np.zeros(MAX_CAPTCHA * CHAR_SET_LEN) def char2pos(c):  try:   k = ord(c)-ord('0')  except:   raise ValueError('No Map')  return k for i, c in enumerate(text):  idx = i * CHAR_SET_LEN + char2pos(c)  vector[idx] = 1 return vector# 向量轉(zhuǎn)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 % CHAR_SET_LEN  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)# 生成一個(gè)訓(xùn)練batchdef get_next_batch(batch_size=128): batch_x = np.zeros([batch_size, IMAGE_HEIGHT * IMAGE_WIDTH]) batch_y = np.zeros([batch_size, MAX_CAPTCHA * CHAR_SET_LEN]) # 有時(shí)生成圖像大小不是(60, 160, 3)  def wrap_gen_captcha_text_and_image():  while True:   text, image = gen_captcha_text_and_image_new()   if image.shape == image_shape:    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為0  batch_y[i, :] = text2vec(text) return batch_x, batch_y####################################################################X = tf.placeholder(tf.float32, [None, IMAGE_HEIGHT * IMAGE_WIDTH])Y = tf.placeholder(tf.float32, [None, MAX_CAPTCHA * CHAR_SET_LEN])keep_prob = tf.placeholder(tf.float32) # dropout# 定義CNNdef crack_captcha_cnn(w_alpha=0.01, b_alpha=0.1): x = tf.reshape(X, shape=[-1, IMAGE_HEIGHT, IMAGE_WIDTH, 1]) # w_c1_alpha = np.sqrt(2.0/(IMAGE_HEIGHT*IMAGE_WIDTH)) # # w_c2_alpha = np.sqrt(2.0/(3*3*32)) # w_c3_alpha = np.sqrt(2.0/(3*3*64)) # w_d1_alpha = np.sqrt(2.0/(8*32*64)) # out_alpha = np.sqrt(2.0/1024) # 定義三層的卷積神經(jīng)網(wǎng)絡(luò) # 定義第一層的卷積神經(jīng)網(wǎng)絡(luò) # 定義第一層權(quán)重 w_c1 = tf.Variable(w_alpha * tf.random_normal([3, 3, 1, 32])) # 定義第一層的偏置 b_c1 = tf.Variable(b_alpha * tf.random_normal([32])) # 定義第一層的激勵(lì)函數(shù) conv1 = tf.nn.relu(tf.nn.bias_add(tf.nn.conv2d(x, w_c1, strides=[1, 1, 1, 1], padding='SAME'), b_c1)) # conv1 為輸入 ksize 表示使用2*2池化，即將2*2的色塊轉(zhuǎn)化成1*1的色塊 conv1 = tf.nn.max_pool(conv1, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME') # dropout防止過擬合。 conv1 = tf.nn.dropout(conv1, keep_prob) # 定義第二層的卷積神經(jīng)網(wǎng)絡(luò) 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) # 定義第三層的卷積神經(jīng)網(wǎng)絡(luò) 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 # 隨機(jī)生成權(quán)重 w_d = tf.Variable(w_alpha * tf.random_normal([1536, 1024])) # 隨機(jī)生成偏置 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 * CHAR_SET_LEN])) b_out = tf.Variable(b_alpha * tf.random_normal([MAX_CAPTCHA * CHAR_SET_LEN])) out = tf.add(tf.matmul(dense, w_out), b_out) # out = tf.nn.softmax(out) return out# 訓(xùn)練def train_crack_captcha_cnn(): # X = tf.placeholder(tf.float32, [None, IMAGE_HEIGHT * IMAGE_WIDTH]) # Y = tf.placeholder(tf.float32, [None, MAX_CAPTCHA * CHAR_SET_LEN]) # keep_prob = tf.placeholder(tf.float32) # dropout output = crack_captcha_cnn() # loss  # loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(output, Y)) loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(output, Y)) # 最后一層用來分類的softmax和sigmoid有什么不同？ # optimizer 為了加快訓(xùn)練 learning_rate應(yīng)該開始大，然后慢慢衰  optimizer = tf.train.AdamOptimizer(learning_rate=0.001).minimize(loss) predict = tf.reshape(output, [-1, MAX_CAPTCHA, CHAR_SET_LEN]) max_idx_p = tf.argmax(predict, 2) max_idx_l = tf.argmax(tf.reshape(Y, [-1, MAX_CAPTCHA, CHAR_SET_LEN]), 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計(jì)算一次準(zhǔn)確率    if step % 100 == 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)     # 如果準(zhǔn)確率大于50%,保存模型,完成訓(xùn)練     if acc > 0.99:      saver.save(sess, "./crack_capcha.model", global_step=step)      break    step += 1## 訓(xùn)練(如果要訓(xùn)練則去掉下面一行的注釋)train_crack_captcha_cnn()def crack_captcha(): output = crack_captcha_cnn() saver = tf.train.Saver() with tf.Session() as sess:  saver.restore(sess, tf.train.latest_checkpoint('.'))  predict = tf.argmax(tf.reshape(output, [-1, MAX_CAPTCHA, CHAR_SET_LEN]), 2)  count = 0  # 因?yàn)闇y試集共40個(gè)...寫的很草率  for i in range(40):   text, image = get_test_captcha_text_and_image(i)   image = convert2gray(image)   captcha_image = image.flatten() / 255   text_list = sess.run(predict, feed_dict={X: [captcha_image], keep_prob: 1})   predict_text = text_list[0].tolist()   predict_text = str(predict_text)   predict_text = predict_text.replace("[", "").replace("]", "").replace(",", "").replace(" ","")   if text == predict_text:    count += 1    check_result = "，預(yù)測結(jié)果正確"   else:    check_result = "，預(yù)測結(jié)果不正確"    print("正確: {} 預(yù)測: {}".format(text, predict_text) + check_result)  print("正確率:" + str(count) + "/40")# 測試(如果要測試則去掉下面一行的注釋)# crack_captcha()

from captcha.image import ImageCaptcha # pip install captchaimport numpy as npfrom PIL import Imageimport random# import matplotlib.pyplot as pltimport osfrom random import choice# 驗(yàn)證碼中的字符, 就不用漢字了number = ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']# alphabet = ['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']# ALPHABET = ['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']root_dir = "d://train"# 驗(yàn)證碼一般都無視大小寫；驗(yàn)證碼長度4個(gè)字符def random_captcha_text(char_set=number, captcha_size=4): captcha_text = [] for i in range(captcha_size):  c = random.choice(char_set)  captcha_text.append(c) return captcha_text# 生成字符對應(yīng)的驗(yàn)證碼def gen_captcha_text_and_image(): image = ImageCaptcha() captcha_text = random_captcha_text() captcha_text = ''.join(captcha_text) captcha = image.generate(captcha_text) # image.write(captcha_text, captcha_text + '.jpg') # 寫到文件 captcha_image = Image.open(captcha) captcha_image = np.array(captcha_image) return captcha_text, captcha_imagedef gen_list(): img_list = [] for parent, dirnames, filenames in os.walk(root_dir): # 三個(gè)參數(shù)：分別返回1.父目錄 2.所有文件夾名字（不含路徑） 3.所有文件名字  for filename in filenames: # 輸出文件信息   img_list.append(filename.replace(".gif",""))   # print("parent is:" + parent)   # print("filename is:" + filename)   # print("the full name of the file is:" + os.path.join(parent, filename)) # 輸出文件路徑信息 return img_listimg_list = gen_list()def gen_captcha_text_and_image_new(): img = choice(img_list) captcha_image = Image.open(root_dir + "//" + img + ".gif") captcha_image = np.array(captcha_image) return img, captcha_image# if __name__ == '__main__':#  # 測試#  # text, image = gen_captcha_text_and_image()#  ##  # f = plt.figure()#  # ax = f.add_subplot(111)#  # ax.text(0.1, 0.9, text, ha='center', va='center', transform=ax.transAxes)#  # plt.imshow(image)#  # plt.show()#  ###  text, image = gen_captcha_text_and_image_new()##  f = plt.figure()#  ax = f.add_subplot(111)#  ax.text(0.1, 0.9, text, ha='center', va='center', transform=ax.transAxes)#  plt.imshow(image)#  plt.show()

from captcha.image import ImageCaptcha # pip install captchaimport numpy as npfrom PIL import Imageimport randomimport matplotlib.pyplot as pltimport osfrom random import choiceroot_dir = "d://test"img_list = []def gen_list(): for parent, dirnames, filenames in os.walk(root_dir): # 三個(gè)參數(shù)：分別返回1.父目錄 2.所有文件夾名字（不含路徑） 3.所有文件名字  for filename in filenames: # 輸出文件信息   img_list.append(filename.replace(".gif",""))   # print("parent is:" + parent)   # print("filename is:" + filename)   # print("the full name of the file is:" + os.path.join(parent, filename)) # 輸出文件路徑信息 return img_listimg_list = gen_list()def get_test_captcha_text_and_image(i=None): img = img_list[i] captcha_image = Image.open(root_dir + "//" + img + ".gif") captcha_image = np.array(captcha_image) return img, captcha_image