您好,登錄后才能下訂單哦!
這篇文章主要為大家展示了“TensorFlow如何實現車牌識別功能”,內容簡而易懂,條理清晰,希望能夠幫助大家解決疑惑,下面讓小編帶領大家一起研究并學習一下“TensorFlow如何實現車牌識別功能”這篇文章吧。
如何使用TensorFlow進行車牌識別,但是,當時采用的數據集是MNIST數字手寫體,只能分類0-9共10個數字,無法分類省份簡稱和字母,局限性較大,無實際意義。
經過圖像定位分割處理,博主收集了相關省份簡稱和26個字母的圖片數據集,結合前述博文中貼出的python+TensorFlow代碼,實現了完整的車牌識別功能。本著分享精神,在此送上全部代碼和車牌數據集。
車牌數據集下載地址(約4000張圖片):tf_car_license_dataset_jb51.rar
省份簡稱訓練+識別代碼(保存文件名為train-license-province.py)(拷貝代碼請務必注意python文本縮進,只要有一處縮進錯誤,就無法得到正確結果,或者出現異常):
#!/usr/bin/python3.5 # -*- coding: utf-8 -*- import sys import os import time import random import numpy as np import tensorflow as tf from PIL import Image SIZE = 1280 WIDTH = 32 HEIGHT = 40 NUM_CLASSES = 6 iterations = 300 SAVER_DIR = "train-saver/province/" PROVINCES = ("京","閩","粵","蘇","滬","浙") nProvinceIndex = 0 time_begin = time.time() # 定義輸入節點,對應于圖片像素值矩陣集合和圖片標簽(即所代表的數字) x = tf.placeholder(tf.float32, shape=[None, SIZE]) y_ = tf.placeholder(tf.float32, shape=[None, NUM_CLASSES]) x_image = tf.reshape(x, [-1, WIDTH, HEIGHT, 1]) # 定義卷積函數 def conv_layer(inputs, W, b, conv_strides, kernel_size, pool_strides, padding): L1_conv = tf.nn.conv2d(inputs, W, strides=conv_strides, padding=padding) L1_relu = tf.nn.relu(L1_conv + b) return tf.nn.max_pool(L1_relu, ksize=kernel_size, strides=pool_strides, padding='SAME') # 定義全連接層函數 def full_connect(inputs, W, b): return tf.nn.relu(tf.matmul(inputs, W) + b) if __name__ =='__main__' and sys.argv[1]=='train': # 第一次遍歷圖片目錄是為了獲取圖片總數 input_count = 0 for i in range(0,NUM_CLASSES): dir = './train_images/training-set/chinese-characters/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: input_count += 1 # 定義對應維數和各維長度的數組 input_images = np.array([[0]*SIZE for i in range(input_count)]) input_labels = np.array([[0]*NUM_CLASSES for i in range(input_count)]) # 第二次遍歷圖片目錄是為了生成圖片數據和標簽 index = 0 for i in range(0,NUM_CLASSES): dir = './train_images/training-set/chinese-characters/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: filename = dir + filename img = Image.open(filename) width = img.size[0] height = img.size[1] for h in range(0, height): for w in range(0, width): # 通過這樣的處理,使數字的線條變細,有利于提高識別準確率 if img.getpixel((w, h)) > 230: input_images[index][w+h*width] = 0 else: input_images[index][w+h*width] = 1 input_labels[index][i] = 1 index += 1 # 第一次遍歷圖片目錄是為了獲取圖片總數 val_count = 0 for i in range(0,NUM_CLASSES): dir = './train_images/validation-set/chinese-characters/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: val_count += 1 # 定義對應維數和各維長度的數組 val_images = np.array([[0]*SIZE for i in range(val_count)]) val_labels = np.array([[0]*NUM_CLASSES for i in range(val_count)]) # 第二次遍歷圖片目錄是為了生成圖片數據和標簽 index = 0 for i in range(0,NUM_CLASSES): dir = './train_images/validation-set/chinese-characters/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: filename = dir + filename img = Image.open(filename) width = img.size[0] height = img.size[1] for h in range(0, height): for w in range(0, width): # 通過這樣的處理,使數字的線條變細,有利于提高識別準確率 if img.getpixel((w, h)) > 230: val_images[index][w+h*width] = 0 else: val_images[index][w+h*width] = 1 val_labels[index][i] = 1 index += 1 with tf.Session() as sess: # 第一個卷積層 W_conv1 = tf.Variable(tf.truncated_normal([8, 8, 1, 16], stddev=0.1), name="W_conv1") b_conv1 = tf.Variable(tf.constant(0.1, shape=[16]), name="b_conv1") conv_strides = [1, 1, 1, 1] kernel_size = [1, 2, 2, 1] pool_strides = [1, 2, 2, 1] L1_pool = conv_layer(x_image, W_conv1, b_conv1, conv_strides, kernel_size, pool_strides, padding='SAME') # 第二個卷積層 W_conv2 = tf.Variable(tf.truncated_normal([5, 5, 16, 32], stddev=0.1), name="W_conv2") b_conv2 = tf.Variable(tf.constant(0.1, shape=[32]), name="b_conv2") conv_strides = [1, 1, 1, 1] kernel_size = [1, 1, 1, 1] pool_strides = [1, 1, 1, 1] L2_pool = conv_layer(L1_pool, W_conv2, b_conv2, conv_strides, kernel_size, pool_strides, padding='SAME') # 全連接層 W_fc1 = tf.Variable(tf.truncated_normal([16 * 20 * 32, 512], stddev=0.1), name="W_fc1") b_fc1 = tf.Variable(tf.constant(0.1, shape=[512]), name="b_fc1") h_pool2_flat = tf.reshape(L2_pool, [-1, 16 * 20*32]) h_fc1 = full_connect(h_pool2_flat, W_fc1, b_fc1) # dropout keep_prob = tf.placeholder(tf.float32) h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob) # readout層 W_fc2 = tf.Variable(tf.truncated_normal([512, NUM_CLASSES], stddev=0.1), name="W_fc2") b_fc2 = tf.Variable(tf.constant(0.1, shape=[NUM_CLASSES]), name="b_fc2") # 定義優化器和訓練op y_conv = tf.matmul(h_fc1_drop, W_fc2) + b_fc2 cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y_, logits=y_conv)) train_step = tf.train.AdamOptimizer((1e-4)).minimize(cross_entropy) correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)) # 初始化saver saver = tf.train.Saver() sess.run(tf.global_variables_initializer()) time_elapsed = time.time() - time_begin print("讀取圖片文件耗費時間:%d秒" % time_elapsed) time_begin = time.time() print ("一共讀取了 %s 個訓練圖像, %s 個標簽" % (input_count, input_count)) # 設置每次訓練op的輸入個數和迭代次數,這里為了支持任意圖片總數,定義了一個余數remainder,譬如,如果每次訓練op的輸入個數為60,圖片總數為150張,則前面兩次各輸入60張,最后一次輸入30張(余數30) batch_size = 60 iterations = iterations batches_count = int(input_count / batch_size) remainder = input_count % batch_size print ("訓練數據集分成 %s 批, 前面每批 %s 個數據,最后一批 %s 個數據" % (batches_count+1, batch_size, remainder)) # 執行訓練迭代 for it in range(iterations): # 這里的關鍵是要把輸入數組轉為np.array for n in range(batches_count): train_step.run(feed_dict={x: input_images[n*batch_size:(n+1)*batch_size], y_: input_labels[n*batch_size:(n+1)*batch_size], keep_prob: 0.5}) if remainder > 0: start_index = batches_count * batch_size; train_step.run(feed_dict={x: input_images[start_index:input_count-1], y_: input_labels[start_index:input_count-1], keep_prob: 0.5}) # 每完成五次迭代,判斷準確度是否已達到100%,達到則退出迭代循環 iterate_accuracy = 0 if it%5 == 0: iterate_accuracy = accuracy.eval(feed_dict={x: val_images, y_: val_labels, keep_prob: 1.0}) print ('第 %d 次訓練迭代: 準確率 %0.5f%%' % (it, iterate_accuracy*100)) if iterate_accuracy >= 0.9999 and it >= 150: break; print ('完成訓練!') time_elapsed = time.time() - time_begin print ("訓練耗費時間:%d秒" % time_elapsed) time_begin = time.time() # 保存訓練結果 if not os.path.exists(SAVER_DIR): print ('不存在訓練數據保存目錄,現在創建保存目錄') os.makedirs(SAVER_DIR) saver_path = saver.save(sess, "%smodel.ckpt"%(SAVER_DIR)) if __name__ =='__main__' and sys.argv[1]=='predict': saver = tf.train.import_meta_graph("%smodel.ckpt.meta"%(SAVER_DIR)) with tf.Session() as sess: model_file=tf.train.latest_checkpoint(SAVER_DIR) saver.restore(sess, model_file) # 第一個卷積層 W_conv1 = sess.graph.get_tensor_by_name("W_conv1:0") b_conv1 = sess.graph.get_tensor_by_name("b_conv1:0") conv_strides = [1, 1, 1, 1] kernel_size = [1, 2, 2, 1] pool_strides = [1, 2, 2, 1] L1_pool = conv_layer(x_image, W_conv1, b_conv1, conv_strides, kernel_size, pool_strides, padding='SAME') # 第二個卷積層 W_conv2 = sess.graph.get_tensor_by_name("W_conv2:0") b_conv2 = sess.graph.get_tensor_by_name("b_conv2:0") conv_strides = [1, 1, 1, 1] kernel_size = [1, 1, 1, 1] pool_strides = [1, 1, 1, 1] L2_pool = conv_layer(L1_pool, W_conv2, b_conv2, conv_strides, kernel_size, pool_strides, padding='SAME') # 全連接層 W_fc1 = sess.graph.get_tensor_by_name("W_fc1:0") b_fc1 = sess.graph.get_tensor_by_name("b_fc1:0") h_pool2_flat = tf.reshape(L2_pool, [-1, 16 * 20*32]) h_fc1 = full_connect(h_pool2_flat, W_fc1, b_fc1) # dropout keep_prob = tf.placeholder(tf.float32) h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob) # readout層 W_fc2 = sess.graph.get_tensor_by_name("W_fc2:0") b_fc2 = sess.graph.get_tensor_by_name("b_fc2:0") # 定義優化器和訓練op conv = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2) + b_fc2) for n in range(1,2): path = "test_images/%s.bmp" % (n) img = Image.open(path) width = img.size[0] height = img.size[1] img_data = [[0]*SIZE for i in range(1)] for h in range(0, height): for w in range(0, width): if img.getpixel((w, h)) < 190: img_data[0][w+h*width] = 1 else: img_data[0][w+h*width] = 0 result = sess.run(conv, feed_dict = {x: np.array(img_data), keep_prob: 1.0}) max1 = 0 max2 = 0 max3 = 0 max1_index = 0 max2_index = 0 max3_index = 0 for j in range(NUM_CLASSES): if result[0][j] > max1: max1 = result[0][j] max1_index = j continue if (result[0][j]>max2) and (result[0][j]<=max1): max2 = result[0][j] max2_index = j continue if (result[0][j]>max3) and (result[0][j]<=max2): max3 = result[0][j] max3_index = j continue nProvinceIndex = max1_index print ("概率: [%s %0.2f%%] [%s %0.2f%%] [%s %0.2f%%]" % (PROVINCES[max1_index],max1*100, PROVINCES[max2_index],max2*100, PROVINCES[max3_index],max3*100)) print ("省份簡稱是: %s" % PROVINCES[nProvinceIndex])
城市代號訓練+識別代碼(保存文件名為train-license-letters.py):
#!/usr/bin/python3.5 # -*- coding: utf-8 -*- import sys import os import time import random import numpy as np import tensorflow as tf from PIL import Image SIZE = 1280 WIDTH = 32 HEIGHT = 40 NUM_CLASSES = 26 iterations = 500 SAVER_DIR = "train-saver/letters/" LETTERS_DIGITS = ("A","B","C","D","E","F","G","H","J","K","L","M","N","P","Q","R","S","T","U","V","W","X","Y","Z","I","O") license_num = "" time_begin = time.time() # 定義輸入節點,對應于圖片像素值矩陣集合和圖片標簽(即所代表的數字) x = tf.placeholder(tf.float32, shape=[None, SIZE]) y_ = tf.placeholder(tf.float32, shape=[None, NUM_CLASSES]) x_image = tf.reshape(x, [-1, WIDTH, HEIGHT, 1]) # 定義卷積函數 def conv_layer(inputs, W, b, conv_strides, kernel_size, pool_strides, padding): L1_conv = tf.nn.conv2d(inputs, W, strides=conv_strides, padding=padding) L1_relu = tf.nn.relu(L1_conv + b) return tf.nn.max_pool(L1_relu, ksize=kernel_size, strides=pool_strides, padding='SAME') # 定義全連接層函數 def full_connect(inputs, W, b): return tf.nn.relu(tf.matmul(inputs, W) + b) if __name__ =='__main__' and sys.argv[1]=='train': # 第一次遍歷圖片目錄是為了獲取圖片總數 input_count = 0 for i in range(0+10,NUM_CLASSES+10): dir = './train_images/training-set/letters/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: input_count += 1 # 定義對應維數和各維長度的數組 input_images = np.array([[0]*SIZE for i in range(input_count)]) input_labels = np.array([[0]*NUM_CLASSES for i in range(input_count)]) # 第二次遍歷圖片目錄是為了生成圖片數據和標簽 index = 0 for i in range(0+10,NUM_CLASSES+10): dir = './train_images/training-set/letters/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: filename = dir + filename img = Image.open(filename) width = img.size[0] height = img.size[1] for h in range(0, height): for w in range(0, width): # 通過這樣的處理,使數字的線條變細,有利于提高識別準確率 if img.getpixel((w, h)) > 230: input_images[index][w+h*width] = 0 else: input_images[index][w+h*width] = 1 #print ("i=%d, index=%d" % (i, index)) input_labels[index][i-10] = 1 index += 1 # 第一次遍歷圖片目錄是為了獲取圖片總數 val_count = 0 for i in range(0+10,NUM_CLASSES+10): dir = './train_images/validation-set/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: val_count += 1 # 定義對應維數和各維長度的數組 val_images = np.array([[0]*SIZE for i in range(val_count)]) val_labels = np.array([[0]*NUM_CLASSES for i in range(val_count)]) # 第二次遍歷圖片目錄是為了生成圖片數據和標簽 index = 0 for i in range(0+10,NUM_CLASSES+10): dir = './train_images/validation-set/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: filename = dir + filename img = Image.open(filename) width = img.size[0] height = img.size[1] for h in range(0, height): for w in range(0, width): # 通過這樣的處理,使數字的線條變細,有利于提高識別準確率 if img.getpixel((w, h)) > 230: val_images[index][w+h*width] = 0 else: val_images[index][w+h*width] = 1 val_labels[index][i-10] = 1 index += 1 with tf.Session() as sess: # 第一個卷積層 W_conv1 = tf.Variable(tf.truncated_normal([8, 8, 1, 16], stddev=0.1), name="W_conv1") b_conv1 = tf.Variable(tf.constant(0.1, shape=[16]), name="b_conv1") conv_strides = [1, 1, 1, 1] kernel_size = [1, 2, 2, 1] pool_strides = [1, 2, 2, 1] L1_pool = conv_layer(x_image, W_conv1, b_conv1, conv_strides, kernel_size, pool_strides, padding='SAME') # 第二個卷積層 W_conv2 = tf.Variable(tf.truncated_normal([5, 5, 16, 32], stddev=0.1), name="W_conv2") b_conv2 = tf.Variable(tf.constant(0.1, shape=[32]), name="b_conv2") conv_strides = [1, 1, 1, 1] kernel_size = [1, 1, 1, 1] pool_strides = [1, 1, 1, 1] L2_pool = conv_layer(L1_pool, W_conv2, b_conv2, conv_strides, kernel_size, pool_strides, padding='SAME') # 全連接層 W_fc1 = tf.Variable(tf.truncated_normal([16 * 20 * 32, 512], stddev=0.1), name="W_fc1") b_fc1 = tf.Variable(tf.constant(0.1, shape=[512]), name="b_fc1") h_pool2_flat = tf.reshape(L2_pool, [-1, 16 * 20*32]) h_fc1 = full_connect(h_pool2_flat, W_fc1, b_fc1) # dropout keep_prob = tf.placeholder(tf.float32) h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob) # readout層 W_fc2 = tf.Variable(tf.truncated_normal([512, NUM_CLASSES], stddev=0.1), name="W_fc2") b_fc2 = tf.Variable(tf.constant(0.1, shape=[NUM_CLASSES]), name="b_fc2") # 定義優化器和訓練op y_conv = tf.matmul(h_fc1_drop, W_fc2) + b_fc2 cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y_, logits=y_conv)) train_step = tf.train.AdamOptimizer((1e-4)).minimize(cross_entropy) correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)) sess.run(tf.global_variables_initializer()) time_elapsed = time.time() - time_begin print("讀取圖片文件耗費時間:%d秒" % time_elapsed) time_begin = time.time() print ("一共讀取了 %s 個訓練圖像, %s 個標簽" % (input_count, input_count)) # 設置每次訓練op的輸入個數和迭代次數,這里為了支持任意圖片總數,定義了一個余數remainder,譬如,如果每次訓練op的輸入個數為60,圖片總數為150張,則前面兩次各輸入60張,最后一次輸入30張(余數30) batch_size = 60 iterations = iterations batches_count = int(input_count / batch_size) remainder = input_count % batch_size print ("訓練數據集分成 %s 批, 前面每批 %s 個數據,最后一批 %s 個數據" % (batches_count+1, batch_size, remainder)) # 執行訓練迭代 for it in range(iterations): # 這里的關鍵是要把輸入數組轉為np.array for n in range(batches_count): train_step.run(feed_dict={x: input_images[n*batch_size:(n+1)*batch_size], y_: input_labels[n*batch_size:(n+1)*batch_size], keep_prob: 0.5}) if remainder > 0: start_index = batches_count * batch_size; train_step.run(feed_dict={x: input_images[start_index:input_count-1], y_: input_labels[start_index:input_count-1], keep_prob: 0.5}) # 每完成五次迭代,判斷準確度是否已達到100%,達到則退出迭代循環 iterate_accuracy = 0 if it%5 == 0: iterate_accuracy = accuracy.eval(feed_dict={x: val_images, y_: val_labels, keep_prob: 1.0}) print ('第 %d 次訓練迭代: 準確率 %0.5f%%' % (it, iterate_accuracy*100)) if iterate_accuracy >= 0.9999 and it >= iterations: break; print ('完成訓練!') time_elapsed = time.time() - time_begin print ("訓練耗費時間:%d秒" % time_elapsed) time_begin = time.time() # 保存訓練結果 if not os.path.exists(SAVER_DIR): print ('不存在訓練數據保存目錄,現在創建保存目錄') os.makedirs(SAVER_DIR) # 初始化saver saver = tf.train.Saver() saver_path = saver.save(sess, "%smodel.ckpt"%(SAVER_DIR)) if __name__ =='__main__' and sys.argv[1]=='predict': saver = tf.train.import_meta_graph("%smodel.ckpt.meta"%(SAVER_DIR)) with tf.Session() as sess: model_file=tf.train.latest_checkpoint(SAVER_DIR) saver.restore(sess, model_file) # 第一個卷積層 W_conv1 = sess.graph.get_tensor_by_name("W_conv1:0") b_conv1 = sess.graph.get_tensor_by_name("b_conv1:0") conv_strides = [1, 1, 1, 1] kernel_size = [1, 2, 2, 1] pool_strides = [1, 2, 2, 1] L1_pool = conv_layer(x_image, W_conv1, b_conv1, conv_strides, kernel_size, pool_strides, padding='SAME') # 第二個卷積層 W_conv2 = sess.graph.get_tensor_by_name("W_conv2:0") b_conv2 = sess.graph.get_tensor_by_name("b_conv2:0") conv_strides = [1, 1, 1, 1] kernel_size = [1, 1, 1, 1] pool_strides = [1, 1, 1, 1] L2_pool = conv_layer(L1_pool, W_conv2, b_conv2, conv_strides, kernel_size, pool_strides, padding='SAME') # 全連接層 W_fc1 = sess.graph.get_tensor_by_name("W_fc1:0") b_fc1 = sess.graph.get_tensor_by_name("b_fc1:0") h_pool2_flat = tf.reshape(L2_pool, [-1, 16 * 20*32]) h_fc1 = full_connect(h_pool2_flat, W_fc1, b_fc1) # dropout keep_prob = tf.placeholder(tf.float32) h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob) # readout層 W_fc2 = sess.graph.get_tensor_by_name("W_fc2:0") b_fc2 = sess.graph.get_tensor_by_name("b_fc2:0") # 定義優化器和訓練op conv = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2) + b_fc2) for n in range(2,3): path = "test_images/%s.bmp" % (n) img = Image.open(path) width = img.size[0] height = img.size[1] img_data = [[0]*SIZE for i in range(1)] for h in range(0, height): for w in range(0, width): if img.getpixel((w, h)) < 190: img_data[0][w+h*width] = 1 else: img_data[0][w+h*width] = 0 result = sess.run(conv, feed_dict = {x: np.array(img_data), keep_prob: 1.0}) max1 = 0 max2 = 0 max3 = 0 max1_index = 0 max2_index = 0 max3_index = 0 for j in range(NUM_CLASSES): if result[0][j] > max1: max1 = result[0][j] max1_index = j continue if (result[0][j]>max2) and (result[0][j]<=max1): max2 = result[0][j] max2_index = j continue if (result[0][j]>max3) and (result[0][j]<=max2): max3 = result[0][j] max3_index = j continue if n == 3: license_num += "-" license_num = license_num + LETTERS_DIGITS[max1_index] print ("概率: [%s %0.2f%%] [%s %0.2f%%] [%s %0.2f%%]" % (LETTERS_DIGITS[max1_index],max1*100, LETTERS_DIGITS[max2_index],max2*100, LETTERS_DIGITS[max3_index],max3*100)) print ("城市代號是: 【%s】" % license_num)
車牌編號訓練+識別代碼(保存文件名為train-license-digits.py):
#!/usr/bin/python3.5 # -*- coding: utf-8 -*- import sys import os import time import random import numpy as np import tensorflow as tf from PIL import Image SIZE = 1280 WIDTH = 32 HEIGHT = 40 NUM_CLASSES = 34 iterations = 1000 SAVER_DIR = "train-saver/digits/" LETTERS_DIGITS = ("0","1","2","3","4","5","6","7","8","9","A","B","C","D","E","F","G","H","J","K","L","M","N","P","Q","R","S","T","U","V","W","X","Y","Z") license_num = "" time_begin = time.time() # 定義輸入節點,對應于圖片像素值矩陣集合和圖片標簽(即所代表的數字) x = tf.placeholder(tf.float32, shape=[None, SIZE]) y_ = tf.placeholder(tf.float32, shape=[None, NUM_CLASSES]) x_image = tf.reshape(x, [-1, WIDTH, HEIGHT, 1]) # 定義卷積函數 def conv_layer(inputs, W, b, conv_strides, kernel_size, pool_strides, padding): L1_conv = tf.nn.conv2d(inputs, W, strides=conv_strides, padding=padding) L1_relu = tf.nn.relu(L1_conv + b) return tf.nn.max_pool(L1_relu, ksize=kernel_size, strides=pool_strides, padding='SAME') # 定義全連接層函數 def full_connect(inputs, W, b): return tf.nn.relu(tf.matmul(inputs, W) + b) if __name__ =='__main__' and sys.argv[1]=='train': # 第一次遍歷圖片目錄是為了獲取圖片總數 input_count = 0 for i in range(0,NUM_CLASSES): dir = './train_images/training-set/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: input_count += 1 # 定義對應維數和各維長度的數組 input_images = np.array([[0]*SIZE for i in range(input_count)]) input_labels = np.array([[0]*NUM_CLASSES for i in range(input_count)]) # 第二次遍歷圖片目錄是為了生成圖片數據和標簽 index = 0 for i in range(0,NUM_CLASSES): dir = './train_images/training-set/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: filename = dir + filename img = Image.open(filename) width = img.size[0] height = img.size[1] for h in range(0, height): for w in range(0, width): # 通過這樣的處理,使數字的線條變細,有利于提高識別準確率 if img.getpixel((w, h)) > 230: input_images[index][w+h*width] = 0 else: input_images[index][w+h*width] = 1 input_labels[index][i] = 1 index += 1 # 第一次遍歷圖片目錄是為了獲取圖片總數 val_count = 0 for i in range(0,NUM_CLASSES): dir = './train_images/validation-set/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: val_count += 1 # 定義對應維數和各維長度的數組 val_images = np.array([[0]*SIZE for i in range(val_count)]) val_labels = np.array([[0]*NUM_CLASSES for i in range(val_count)]) # 第二次遍歷圖片目錄是為了生成圖片數據和標簽 index = 0 for i in range(0,NUM_CLASSES): dir = './train_images/validation-set/%s/' % i # 這里可以改成你自己的圖片目錄,i為分類標簽 for rt, dirs, files in os.walk(dir): for filename in files: filename = dir + filename img = Image.open(filename) width = img.size[0] height = img.size[1] for h in range(0, height): for w in range(0, width): # 通過這樣的處理,使數字的線條變細,有利于提高識別準確率 if img.getpixel((w, h)) > 230: val_images[index][w+h*width] = 0 else: val_images[index][w+h*width] = 1 val_labels[index][i] = 1 index += 1 with tf.Session() as sess: # 第一個卷積層 W_conv1 = tf.Variable(tf.truncated_normal([8, 8, 1, 16], stddev=0.1), name="W_conv1") b_conv1 = tf.Variable(tf.constant(0.1, shape=[16]), name="b_conv1") conv_strides = [1, 1, 1, 1] kernel_size = [1, 2, 2, 1] pool_strides = [1, 2, 2, 1] L1_pool = conv_layer(x_image, W_conv1, b_conv1, conv_strides, kernel_size, pool_strides, padding='SAME') # 第二個卷積層 W_conv2 = tf.Variable(tf.truncated_normal([5, 5, 16, 32], stddev=0.1), name="W_conv2") b_conv2 = tf.Variable(tf.constant(0.1, shape=[32]), name="b_conv2") conv_strides = [1, 1, 1, 1] kernel_size = [1, 1, 1, 1] pool_strides = [1, 1, 1, 1] L2_pool = conv_layer(L1_pool, W_conv2, b_conv2, conv_strides, kernel_size, pool_strides, padding='SAME') # 全連接層 W_fc1 = tf.Variable(tf.truncated_normal([16 * 20 * 32, 512], stddev=0.1), name="W_fc1") b_fc1 = tf.Variable(tf.constant(0.1, shape=[512]), name="b_fc1") h_pool2_flat = tf.reshape(L2_pool, [-1, 16 * 20*32]) h_fc1 = full_connect(h_pool2_flat, W_fc1, b_fc1) # dropout keep_prob = tf.placeholder(tf.float32) h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob) # readout層 W_fc2 = tf.Variable(tf.truncated_normal([512, NUM_CLASSES], stddev=0.1), name="W_fc2") b_fc2 = tf.Variable(tf.constant(0.1, shape=[NUM_CLASSES]), name="b_fc2") # 定義優化器和訓練op y_conv = tf.matmul(h_fc1_drop, W_fc2) + b_fc2 cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y_, logits=y_conv)) train_step = tf.train.AdamOptimizer((1e-4)).minimize(cross_entropy) correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)) sess.run(tf.global_variables_initializer()) time_elapsed = time.time() - time_begin print("讀取圖片文件耗費時間:%d秒" % time_elapsed) time_begin = time.time() print ("一共讀取了 %s 個訓練圖像, %s 個標簽" % (input_count, input_count)) # 設置每次訓練op的輸入個數和迭代次數,這里為了支持任意圖片總數,定義了一個余數remainder,譬如,如果每次訓練op的輸入個數為60,圖片總數為150張,則前面兩次各輸入60張,最后一次輸入30張(余數30) batch_size = 60 iterations = iterations batches_count = int(input_count / batch_size) remainder = input_count % batch_size print ("訓練數據集分成 %s 批, 前面每批 %s 個數據,最后一批 %s 個數據" % (batches_count+1, batch_size, remainder)) # 執行訓練迭代 for it in range(iterations): # 這里的關鍵是要把輸入數組轉為np.array for n in range(batches_count): train_step.run(feed_dict={x: input_images[n*batch_size:(n+1)*batch_size], y_: input_labels[n*batch_size:(n+1)*batch_size], keep_prob: 0.5}) if remainder > 0: start_index = batches_count * batch_size; train_step.run(feed_dict={x: input_images[start_index:input_count-1], y_: input_labels[start_index:input_count-1], keep_prob: 0.5}) # 每完成五次迭代,判斷準確度是否已達到100%,達到則退出迭代循環 iterate_accuracy = 0 if it%5 == 0: iterate_accuracy = accuracy.eval(feed_dict={x: val_images, y_: val_labels, keep_prob: 1.0}) print ('第 %d 次訓練迭代: 準確率 %0.5f%%' % (it, iterate_accuracy*100)) if iterate_accuracy >= 0.9999 and it >= iterations: break; print ('完成訓練!') time_elapsed = time.time() - time_begin print ("訓練耗費時間:%d秒" % time_elapsed) time_begin = time.time() # 保存訓練結果 if not os.path.exists(SAVER_DIR): print ('不存在訓練數據保存目錄,現在創建保存目錄') os.makedirs(SAVER_DIR) # 初始化saver saver = tf.train.Saver() saver_path = saver.save(sess, "%smodel.ckpt"%(SAVER_DIR)) if __name__ =='__main__' and sys.argv[1]=='predict': saver = tf.train.import_meta_graph("%smodel.ckpt.meta"%(SAVER_DIR)) with tf.Session() as sess: model_file=tf.train.latest_checkpoint(SAVER_DIR) saver.restore(sess, model_file) # 第一個卷積層 W_conv1 = sess.graph.get_tensor_by_name("W_conv1:0") b_conv1 = sess.graph.get_tensor_by_name("b_conv1:0") conv_strides = [1, 1, 1, 1] kernel_size = [1, 2, 2, 1] pool_strides = [1, 2, 2, 1] L1_pool = conv_layer(x_image, W_conv1, b_conv1, conv_strides, kernel_size, pool_strides, padding='SAME') # 第二個卷積層 W_conv2 = sess.graph.get_tensor_by_name("W_conv2:0") b_conv2 = sess.graph.get_tensor_by_name("b_conv2:0") conv_strides = [1, 1, 1, 1] kernel_size = [1, 1, 1, 1] pool_strides = [1, 1, 1, 1] L2_pool = conv_layer(L1_pool, W_conv2, b_conv2, conv_strides, kernel_size, pool_strides, padding='SAME') # 全連接層 W_fc1 = sess.graph.get_tensor_by_name("W_fc1:0") b_fc1 = sess.graph.get_tensor_by_name("b_fc1:0") h_pool2_flat = tf.reshape(L2_pool, [-1, 16 * 20*32]) h_fc1 = full_connect(h_pool2_flat, W_fc1, b_fc1) # dropout keep_prob = tf.placeholder(tf.float32) h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob) # readout層 W_fc2 = sess.graph.get_tensor_by_name("W_fc2:0") b_fc2 = sess.graph.get_tensor_by_name("b_fc2:0") # 定義優化器和訓練op conv = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2) + b_fc2) for n in range(3,8): path = "test_images/%s.bmp" % (n) img = Image.open(path) width = img.size[0] height = img.size[1] img_data = [[0]*SIZE for i in range(1)] for h in range(0, height): for w in range(0, width): if img.getpixel((w, h)) < 190: img_data[0][w+h*width] = 1 else: img_data[0][w+h*width] = 0 result = sess.run(conv, feed_dict = {x: np.array(img_data), keep_prob: 1.0}) max1 = 0 max2 = 0 max3 = 0 max1_index = 0 max2_index = 0 max3_index = 0 for j in range(NUM_CLASSES): if result[0][j] > max1: max1 = result[0][j] max1_index = j continue if (result[0][j]>max2) and (result[0][j]<=max1): max2 = result[0][j] max2_index = j continue if (result[0][j]>max3) and (result[0][j]<=max2): max3 = result[0][j] max3_index = j continue license_num = license_num + LETTERS_DIGITS[max1_index] print ("概率: [%s %0.2f%%] [%s %0.2f%%] [%s %0.2f%%]" % (LETTERS_DIGITS[max1_index],max1*100, LETTERS_DIGITS[max2_index],max2*100, LETTERS_DIGITS[max3_index],max3*100)) print ("車牌編號是: 【%s】" % license_num)
保存好上面三個python腳本后,我們首先進行省份簡稱訓練。在運行代碼之前,需要先把數據集解壓到訓練腳本所在目錄。然后,在命令行中進入腳本所在目錄,輸入執行如下命令:
python train-license-province.py train
訓練結果如下:
然后進行省份簡稱識別,在命令行輸入執行如下命令:
python train-license-province.py predict
執行城市代號訓練(相當于訓練26個字母):
python train-license-letters.py train
識別城市代號:
python train-license-letters.py predict
執行車牌編號訓練(相當于訓練24個字母+10個數字,我國交通法規規定車牌編號中不包含字母I和O):
python train-license-digits.py train
識別車牌編號:
python train-license-digits.py predict
可以看到,在測試圖片上,識別準確率很高。識別結果是閩O-1672Q。
下圖是測試圖片的車牌原圖:
以上是“TensorFlow如何實現車牌識別功能”這篇文章的所有內容,感謝各位的閱讀!相信大家都有了一定的了解,希望分享的內容對大家有所幫助,如果還想學習更多知識,歡迎關注億速云行業資訊頻道!
免責聲明:本站發布的內容(圖片、視頻和文字)以原創、轉載和分享為主,文章觀點不代表本網站立場,如果涉及侵權請聯系站長郵箱:is@yisu.com進行舉報,并提供相關證據,一經查實,將立刻刪除涉嫌侵權內容。