訓練“她”，讓“她”認識我！
學習自@little_wang
原文地址：

http://www.itdecent.cn/p/38ca6daf6b40

1.獲取我的人臉數(shù)據(jù)：

注意以下有一個引用庫import cv2
為順利找到依賴庫函數(shù)，需要先安裝庫：

pip install opencv-python

• 使用opencv打開攝像頭，獲取人臉
• 對圖像做一些預處理，如處理成64*64大小的圖片
• 獲取期間，做一些明暗處理，以增加圖像的噪聲干擾，使得訓練出來的模型具備一定的泛化能力
• 共獲取200張照片

#!/usr/bin/python
#coding=utf-8
''' face detect
https://github.com/seathiefwang/FaceRecognition-tensorflow
http://tumumu.cn/2017/05/02/deep-learning-face/
'''
# pylint: disable=invalid-name
import os
import random
import numpy as np
import cv2

def createdir(*args):
    ''' create dir'''
    for item in args:
        if not os.path.exists(item):
            os.makedirs(item)

IMGSIZE = 64


def getpaddingSize(shape):
    ''' get size to make image to be a square rect '''
    h, w = shape
    longest = max(h, w)
    result = (np.array([longest]*4, int) - np.array([h, h, w, w], int)) // 2
    return result.tolist()

def dealwithimage(img, h=64, w=64):
    ''' dealwithimage '''
    #img = cv2.imread(imgpath)
    top, bottom, left, right = getpaddingSize(img.shape[0:2])
    img = cv2.copyMakeBorder(img, top, bottom, left, right, cv2.BORDER_CONSTANT, value=[0, 0, 0])
    img = cv2.resize(img, (h, w))
    return img

def relight(imgsrc, alpha=1, bias=0):
    '''relight'''
    imgsrc = imgsrc.astype(float)
    imgsrc = imgsrc * alpha + bias
    imgsrc[imgsrc < 0] = 0
    imgsrc[imgsrc > 255] = 255
    imgsrc = imgsrc.astype(np.uint8)
    return imgsrc

def getfacefromcamera(outdir):
    createdir(outdir)
    camera = cv2.VideoCapture(0)
    haar = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
    n = 1
    while 1:
        if (n <= 200):
            print('It`s processing %s image.' % n)
            # 讀幀
            success, img = camera.read()

            gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
            faces = haar.detectMultiScale(gray_img, 1.3, 5)
            for f_x, f_y, f_w, f_h in faces:
                face = img[f_y:f_y+f_h, f_x:f_x+f_w]
                face = cv2.resize(face, (IMGSIZE, IMGSIZE))
                #could deal with face to train
                face = relight(face, random.uniform(0.5, 1.5), random.randint(-50, 50))
                cv2.imwrite(os.path.join(outdir, str(n)+'.jpg'), face)

                cv2.putText(img, 'haha', (f_x, f_y - 20), cv2.FONT_HERSHEY_SIMPLEX, 1, 255, 2)  #顯示名字
                img = cv2.rectangle(img, (f_x, f_y), (f_x + f_w, f_y + f_h), (255, 0, 0), 2)
                n+=1
            cv2.imshow('img', img)
            key = cv2.waitKey(30) & 0xff
            if key == 27:
                break
        else:
            break
    camera.release()
    cv2.destroyAllWindows()

if __name__ == '__main__':
    name = input('please input yourename: ')
    getfacefromcamera(os.path.join('./image/trainfaces', name))

執(zhí)行完以后效果是這樣的（原諒我作了處理（-——-））

Inked捕獲_LI.jpg

2.創(chuàng)建CNN網(wǎng)絡：

#!/usr/bin/python
#coding=utf-8
''' face detect convolution'''
# pylint: disable=invalid-name
import os
import logging as log
import matplotlib.pyplot as plt
import common
import numpy as np
from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf
import cv2

SIZE = 64
x_data = tf.placeholder(tf.float32, [None, SIZE, SIZE, 3])
y_data = tf.placeholder(tf.float32, [None, None])

keep_prob_5 = tf.placeholder(tf.float32)
keep_prob_75 = tf.placeholder(tf.float32)

def weightVariable(shape):
    ''' build weight variable'''
    init = tf.random_normal(shape, stddev=0.01)
    #init = tf.truncated_normal(shape, stddev=0.01)
    return tf.Variable(init)

def biasVariable(shape):
    ''' build bias variable'''
    init = tf.random_normal(shape)
    #init = tf.truncated_normal(shape, stddev=0.01)
    return tf.Variable(init)

def conv2d(x, W):
    ''' conv2d by 1, 1, 1, 1'''
    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')

def maxPool(x):
    ''' max pooling'''
    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')

def dropout(x, keep):
    ''' drop out'''
    return tf.nn.dropout(x, keep)

def cnnLayer(classnum):
    ''' create cnn layer'''
    # 第一層
    W1 = weightVariable([3, 3, 3, 32]) # 卷積核大小(3,3)， 輸入通道(3)， 輸出通道(32)
    b1 = biasVariable([32])
    conv1 = tf.nn.relu(conv2d(x_data, W1) + b1)
    pool1 = maxPool(conv1)
    # 減少過擬合，隨機讓某些權(quán)重不更新
    drop1 = dropout(pool1, keep_prob_5) # 32 * 32 * 32 多個輸入channel 被filter內(nèi)積掉了

    # 第二層
    W2 = weightVariable([3, 3, 32, 64])
    b2 = biasVariable([64])
    conv2 = tf.nn.relu(conv2d(drop1, W2) + b2)
    pool2 = maxPool(conv2)
    drop2 = dropout(pool2, keep_prob_5) # 64 * 16 * 16

    # 第三層
    W3 = weightVariable([3, 3, 64, 64])
    b3 = biasVariable([64])
    conv3 = tf.nn.relu(conv2d(drop2, W3) + b3)
    pool3 = maxPool(conv3)
    drop3 = dropout(pool3, keep_prob_5) # 64 * 8 * 8

    # 全連接層
    Wf = weightVariable([8*16*32, 512])
    bf = biasVariable([512])
    drop3_flat = tf.reshape(drop3, [-1, 8*16*32])
    dense = tf.nn.relu(tf.matmul(drop3_flat, Wf) + bf)
    dropf = dropout(dense, keep_prob_75)

    # 輸出層
    Wout = weightVariable([512, classnum])
    bout = weightVariable([classnum])
    #out = tf.matmul(dropf, Wout) + bout
    out = tf.add(tf.matmul(dropf, Wout), bout)
    return out

def train(train_x, train_y, tfsavepath):
    ''' train'''
    log.debug('train')
    out = cnnLayer(train_y.shape[1])
    cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=out, labels=y_data))
    train_step = tf.train.AdamOptimizer(0.01).minimize(cross_entropy)
    accuracy = tf.reduce_mean(tf.cast(tf.equal(tf.argmax(out, 1), tf.argmax(y_data, 1)), tf.float32))

    saver = tf.train.Saver()
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        batch_size = 10
        num_batch = len(train_x) // 10
        for n in range(10):
            r = np.random.permutation(len(train_x))
            train_x = train_x[r, :]
            train_y = train_y[r, :]

            for i in range(num_batch):
                batch_x = train_x[i*batch_size : (i+1)*batch_size]
                batch_y = train_y[i*batch_size : (i+1)*batch_size]
                _, loss = sess.run([train_step, cross_entropy],\
                                   feed_dict={x_data:batch_x, y_data:batch_y,
                                              keep_prob_5:0.75, keep_prob_75:0.75})

                print(n*num_batch+i, loss)

        # 獲取測試數(shù)據(jù)的準確率
        acc = accuracy.eval({x_data:train_x, y_data:train_y, keep_prob_5:1.0, keep_prob_75:1.0})
        print('after 10 times run: accuracy is ', acc)
        saver.save(sess, tfsavepath)

def validate(test_x, tfsavepath):
    ''' validate '''
    output = cnnLayer(2)
    #predict = tf.equal(tf.argmax(output, 1), tf.argmax(y_data, 1))
    predict = output

    saver = tf.train.Saver()
    with tf.Session() as sess:
        #sess.run(tf.global_variables_initializer())
        saver.restore(sess, tfsavepath)
        res = sess.run([predict, tf.argmax(output, 1)],
                       feed_dict={x_data: test_x,
                                  keep_prob_5:1.0, keep_prob_75: 1.0})
        return res

if __name__ == '__main__':
    pass

使用tf創(chuàng)建3層cnn，3 * 3的filter，輸入為rgb所以：

• 第一層的channel是3，圖像寬高為64，輸出32個filter，maxpooling是縮放一倍
• 第二層的輸入為32個channel，寬高是32，輸出為64個filter，maxpooling是縮放一倍
• 第三層的輸入為64個channel，寬高是16，輸出為64個filter，maxpooling是縮放一倍
  所以最后輸入的圖像是8 * 8 * 64，卷積層和全連接層都設置了dropout參數(shù)

將輸入的8 * 8 * 64的多維度，進行flatten，映射到512個數(shù)據(jù)上，然后進行softmax，輸出到onehot類別上，類別的輸入根據(jù)采集的人員的個數(shù)來確定。

3.識別人臉分類

def getfileandlabel(filedir):
    ''' get path and host paire and class index to name'''
    dictdir = dict([[name, os.path.join(filedir, name)] \
                    for name in os.listdir(filedir) if os.path.isdir(os.path.join(filedir, name))])
                    #for (path, dirnames, _) in os.walk(filedir) for dirname in dirnames])

    dirnamelist, dirpathlist = dictdir.keys(), dictdir.values()
    indexlist = list(range(len(dirnamelist)))

    return list(zip(dirpathlist, onehot(indexlist))), dict(zip(indexlist, dirnamelist))
    
pathlabelpair, indextoname = getfileandlabel('./image/trainfaces')
train_x, train_y = readimage(pathlabelpair)
train_x = train_x.astype(np.float32) / 255.0
myconv.train(train_x, train_y, savepath)

• 將人臉從子目錄內(nèi)讀出來，根據(jù)不同的人名，分配不同的onehot值，這里是按照遍歷的順序分配序號，然后訓練，完成之后會保存checkpoint
• 圖像識別之前將像素值轉(zhuǎn)換為0到1的范圍
• 需要多次訓練的話，把checkpoint下面的上次訓練結(jié)果刪除，代碼有個判斷，有上一次的訓練結(jié)果，就不會再訓練了

4.識別圖像

def testfromcamera(chkpoint):
    camera = cv2.VideoCapture(0)
    haar = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
    pathlabelpair, indextoname = getfileandlabel('./image/trainfaces')
    output = myconv.cnnLayer(len(pathlabelpair))
    #predict = tf.equal(tf.argmax(output, 1), tf.argmax(y_data, 1))
    predict = output

    saver = tf.train.Saver()
    with tf.Session() as sess:
        #sess.run(tf.global_variables_initializer())
        saver.restore(sess, chkpoint)
        
        n = 1
        while 1:
            if (n <= 20000):
                print('It`s processing %s image.' % n)
                # 讀幀
                success, img = camera.read()

                gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
                faces = haar.detectMultiScale(gray_img, 1.3, 5)
                for f_x, f_y, f_w, f_h in faces:
                    face = img[f_y:f_y+f_h, f_x:f_x+f_w]
                    face = cv2.resize(face, (IMGSIZE, IMGSIZE))
                    #could deal with face to train
                    test_x = np.array([face])
                    test_x = test_x.astype(np.float32) / 255.0
                    
                    res = sess.run([predict, tf.argmax(output, 1)],\
                                   feed_dict={myconv.x_data: test_x,\
                                   myconv.keep_prob_5:1.0, myconv.keep_prob_75: 1.0})
                    print(res)

                    cv2.putText(img, indextoname[res[1][0]], (f_x, f_y - 20), cv2.FONT_HERSHEY_SIMPLEX, 1, 255, 2)  #顯示名字
                    img = cv2.rectangle(img, (f_x, f_y), (f_x + f_w, f_y + f_h), (255, 0, 0), 2)
                    n+=1
                cv2.imshow('img', img)
                key = cv2.waitKey(30) & 0xff
                if key == 27:
                    break
            else:
                break
    camera.release()
    cv2.destroyAllWindows()

• 從訓練的結(jié)果中恢復訓練識別的參數(shù)，然后用于新的識別判斷
• 打開攝像頭，采集到圖片之后，進行人臉檢測，檢測出來之后，進行人臉識別，根據(jù)結(jié)果對應到人員名字，顯示在圖片中人臉的上面

5.測試效果如下

測試結(jié)果.PNG