opencv實(shí)現(xiàn)人臉檢測(cè),tensorflow利用cnn實(shí)現(xiàn)人臉識(shí)別,python完成
github地址: https://github.com/wangdxh/tensorflow-learn
基礎(chǔ)知識(shí)
- 微積分求導(dǎo),求偏導(dǎo)(線代吳恩達(dá)的課程會(huì)介紹)
- 吳恩達(dá) 機(jī)器學(xué)習(xí) b站鏈接
- 臺(tái)大李宏毅 b站鏈接
- tensorflow Building Machine Learning Projects with TensorFlow paswd:nasm
獲得人臉數(shù)據(jù)
tensorflow_face_camera.py
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()
name = input('please input yourename: ')
getfacefromcamera(os.path.join('./image/trainfaces', name))
- 根據(jù)輸入的名字在./image/trainfaces目錄下面創(chuàng)建子目錄,將本次采集的頭像保存在該目錄之下
- 使用opencv打開攝像頭,獲取頭像
- 檢測(cè)出人臉的區(qū)域,調(diào)整一下亮暗度,將圖片保存
- 保存200張之后,采集結(jié)束
創(chuàng)建cnn網(wǎng)絡(luò)
具體在tensorflow_face_conv.py
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(jī)讓某些權(quán)重不更新
drop1 = dropout(pool1, keep_prob_5) # 32 * 32 * 32 多個(gè)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
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(gè)filter,maxpooling是縮放一倍
- 第二層的輸入為32個(gè)channel,寬高是32,輸出為64個(gè)filter,maxpooling是縮放一倍
- 第三層的輸入為64個(gè)channel,寬高是16,輸出為64個(gè)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(shè)置了dropout參數(shù)
將輸入的8 * 8 * 64的多維度,進(jìn)行flatten,映射到512個(gè)數(shù)據(jù)上,然后進(jìn)行softmax,輸出到onehot類別上,類別的輸入根據(jù)采集的人員的個(gè)數(shù)來確定。
圖片發(fā)自簡(jiǎn)書App
識(shí)別人臉分類
tensorflow_face.py
訓(xùn)練神經(jīng)網(wǎng)絡(luò)
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(hào),然后訓(xùn)練,完成之后會(huì)保存checkpoint
- 圖像識(shí)別之前將像素值轉(zhuǎn)換為0到1的范圍
- 需要多次訓(xùn)練的話,把checkpoint下面的上次訓(xùn)練結(jié)果刪除,代碼有個(gè)判斷,有上一次的訓(xùn)練結(jié)果,就不會(huì)再訓(xùn)練了
識(shí)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(xùn)練的結(jié)果中恢復(fù)訓(xùn)練識(shí)別的參數(shù),然后用于新的識(shí)別判斷
- 打開攝像頭,采集到圖片之后,進(jìn)行人臉檢測(cè),檢測(cè)出來之后,進(jìn)行人臉識(shí)別,根據(jù)結(jié)果對(duì)應(yīng)到人員名字,顯示在圖片中人臉的上面
遺留問題
weight 和 bias 的初始化好像有些問題,隨機(jī)初始化會(huì)造成在某些情況下cost很大,梯度下不去,導(dǎo)致train結(jié)果很差。重新跑一次命中率又搞了,隨機(jī)這里可以使用truncated_normal再測(cè)試測(cè)試。
輸出的種類數(shù)目是根據(jù)采集的人數(shù)去動(dòng)態(tài)變化,但是沒有給陌生人預(yù)留class,所以結(jié)果肯定在某個(gè)采集的人中,區(qū)別不出陌生人來,可以在onehot的個(gè)數(shù)加1,增加一個(gè)陌生人類別,再進(jìn)行測(cè)試。