用CNN卷积神经网络进行人脸识别
1：对数据的处理class类
# -*- coding: utf-8 -*-
import os
import sys
import numpy as np
import cv2
IMAGE_SIZE = 64
#按照指定图像大小调整尺寸
def resize_image(image, height = IMAGE_SIZE, width = IMAGE_SIZE):
top, bottom, left, right = (0, 0, 0, 0)
h, w, _ = image.shape
#对于长宽不相等的图片，找到最长的一边
longest_edge = max(h, w)
#计算短边需要增加多上像素宽度使其与长边等长
if h < longest_edge:
dh = longest_edge - h
top = dh // 2
bottom = dh - top
elif w < longest_edge:
dw = longest_edge - w
left = dw // 2
right = dw - left
else:
pass
#RGB颜色
BLACK = [0, 0, 0]
#给图像增加边界，是图片长、宽等长，cv2.BORDER_CONSTANT指定边界颜色由value指定
constant = cv2.copyMakeBorder(image, top , bottom, left, right, cv2.BORDER_CONSTANT, value = BLACK)
#调整图像大小并返回
return cv2.resize(constant, (height, width))
#读取训练数据
images = []
labels = []
def read_path(path_name):
for dir_item in os.listdir(path_name):
#从初始路径开始叠加，合并成可识别的操作路径
full_path = os.path.abspath(os.path.join(path_name, dir_item))
if os.path.isdir(full_path):
read_path(full_path)
else:
if dir_item.endswith('.jpg'):
image = cv2.imread(full_path)
image = resize_image(image, IMAGE_SIZE, IMAGE_SIZE)
#cv2.imwrite('1.jpg', image)
images.append(image)
#print(full_path.split("\\")[3])
labels.append(full_path.split("\\")[3])
return images,labels
#从指定路径读取训练数据
def load_dataset(path_name):
images,labels = read_path(path_name)
#将输入的所有图片转成四维数组，尺寸为(图片数量*IMAGE_SIZE*IMAGE_SIZE*3)
#两个人共1200张图片，IMAGE_SIZE为64，故对我来说尺寸为1200 * 64 * 64 * 3
#图片为64 * 64像素,一个像素3个颜色值(RGB)
images = np.array(images)
#标注数据，'me'文件夹下都是我的脸部图像，全部指定为0，另外一个文件夹下是闺女的，全部指定为1
labels = np.array([0 if label.startswith('meng') else 1 for label in labels])
return images, labels
2：对model的训练，将训练好的model进行保存
#CNN网络模型类
import random
from keras.preprocessing.image import ImageDataGenerator
from keras.layers import Conv2D,MaxPooling2D,Activation,Dropout,Flatten,Dense,Convolution2D
from sklearn.model_selection import train_test_split
from keras.models import Sequential
from keras.optimizers import SGD
from keras.utils import np_utils
from keras.models import load_model
from load_face_dataset import load_dataset, resize_image, IMAGE_SIZE
from keras import backend as k
class Model:
def __init__(self):
self.model = None
#建立模型
def build_model(self, dataset, nb_classes = 2):
#构建一个空的网络模型，它是一个线性堆叠模型，各神经网络层会被顺序添加，专业名称为序贯模型或线性堆叠模型
self.model = Sequential()
self.modelmodel.add(Conv2D(
filters=32,
activation='relu',
padding='same',
input_shape=(IMAGE_SIZE, IMAGE_SIZE, img_channels),
strides=(1, 1),
name='Con2D1',
kernel_size=(3, 3)
))
self.model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
self.model.add(Conv2D(
filters=64,
strides=(1, 1),
padding='same',
kernel_size=(3, 3),
name='Con2D2',
activation='relu'
))
self.model.add(MaxPooling2D(pool_size=(2, 2)))
self.model.add(Dropout(0.5))
self.model.add(Flatten())
self.model.add(Dense(128, activation='relu'))
self.model.add(Dense(64, activation='relu'))
self.model.add(Dense(32, activation='relu'))
self.model.add(Dense(16, activation='relu'))
self.model.add(Dense(8, activation='relu'))
self.model.add(Dense(2, activation='softmax'))
self.model.compile(loss='categorical_crossentropy', optimizer='adagrad', metrics=['accuracy'])
# 最后定义损失函数
self.model.fit(self.train_images, self.train_labels, validation_data=(self.valid_inamges, self.valid_labels),
epochs=40, batch_size=25)
# 最后放入批量样本，进行训练
scores = self.model.evaluate(self.test_images, self.test_labels, verbose=0)
self.model.save('model_weight.h5')
# 识别人脸
def face_predict(self, image):
# 依然是根据后端系统确定维度顺序
if k.image_dim_ordering() == 'th' and image.shape != (1, 3, IMAGE_SIZE, IMAGE_SIZE):
image = resize_image(image)
image = image.reshape((1, 3, IMAGE_SIZE, IMAGE_SIZE))
elif k.image_dim_ordering() == 'tf' and image.shape != (1, IMAGE_SIZE, IMAGE_SIZE, 3):
image = resize_image(image)
image = image.reshape((1, IMAGE_SIZE, IMAGE_SIZE, 3))
# 浮点并归一化
image = image.astype('float32')
image /= 255.0
# 给出输入属于各个类别的概率，这里是二值类别，输出属于0和1的概率各为多少
# 给出类别预测：0或者1
result = self.model.predict_classes(image)
print("result----"+ str(result))
# 返回类别预测结果
return result[0]
# 训练模型
# 训练模型
def train(self, dataset, batch_size=25, nb_epoch=10, data_augmentation=True):
sgd = SGD(lr=0.01, decay=1e-6,
momentum=0.9, nesterov=True)
self.model.compile(loss='categorical_crossentropy',
optimizer=sgd,
metrics=['accuracy'])
# 不使用数据提升，所谓的提升就是从我们提供的训练数据中利用旋转、翻转、加噪声等方法创造新的
# 训练数据，有意识的提升训练数据规模，增加模型训练量
if not data_augmentation:
self.model.fit(dataset.train_images,
dataset.train_labels,
batch_size=batch_size,
nb_epoch=nb_epoch,
validation_data=(dataset.valid_images, dataset.valid_labels),
shuffle=True)
else:
datagen = ImageDataGenerator(
featurewise_center=False,
samplewise_center=False,
featurewise_std_normalization=False,
samplewise_std_normalization=False,
zca_whitening=False,
rotation_range=20,
width_shift_range=0.2,
height_shift_range=0.2,
horizontal_flip=True,
vertical_flip=False)
# 计算整个训练样本集的数量以用于特征值归一化、ZCA白化等处理
datagen.fit(dataset.train_images)
# 利用生成器开始训练模型
self.model.fit_generator(datagen.flow(dataset.train_images, dataset.train_labels, batch_size=batch_size),
samples_per_epoch=dataset.train_images.shape[0], nb_epoch=nb_epoch,
validation_data=(dataset.valid_images, dataset.valid_labels))
MODEL_PATH = 'me.face.model.h5'
def save_model(self, file_path=MODEL_PATH):
self.model.save(file_path)
def load_model(self, file_path=MODEL_PATH):
self.model = load_model(file_path)
def evaluate(self, dataset):
score = self.model.evaluate(dataset.test_images, dataset.test_labels, verbose=1)
print("%s: %.2f%%" % (self.model.metrics_names[1], score[1] * 100))
class Dataset:
def __init__(self,path_name):
self.train_images = None
self.train_labels = None
self.valid_inamges = None
self.valid_labels = None
self.test_images = None
self.test_labels = None
self.path_name = path_name
self.input_shape = None
# 加载数据集并按照交叉验证的原则划分数据集并进行相关预处理工作
def load(self,img_rows = IMAGE_SIZE,img_cols = IMAGE_SIZE,img_channels = 3 ,nb_classes = 2):
#加载数据到内存
images,labels = load_dataset(self.path_name)
train_images,valid_images,train_labels,valid_labels = \
train_test_split(images,labels,test_size=0.3,random_state=random.randint(0,100))
_,test_images,_,test_labels = train_test_split(images,labels,test_size=0.5,random_state=random.randint(0,100))
# 当前的维度顺序如果为'th'，则输入图片数据时的顺序为：channels,rows,cols，否则:rows,cols,channels
# 这部分代码就是根据keras库要求的维度顺序重组训练数据集
if k.image_dim_ordering() == 'th':
train_images = train_images.reshape(train_images.shape[0], img_channels, img_rows, img_cols)
valid_images = valid_images.reshape(valid_images.shape[0], img_channels, img_rows, img_cols)
test_images = test_images.reshape(test_images.shape[0], img_channels, img_rows, img_cols)
self.input_shape = (img_channels, img_rows, img_cols)
else:
train_images = train_images.reshape(train_images.shape[0], img_rows, img_cols, img_channels)
valid_images = valid_images.reshape(valid_images.shape[0], img_rows, img_cols, img_channels)
test_images = test_images.reshape(test_images.shape[0], img_rows, img_cols, img_channels)
self.input_shape = (img_rows, img_cols, img_channels)
# 我们的模型使用categorical_crossentropy作为损失函数，因此需要根据类别数量nb_classes将
# 类别标签进行one-hot编码使其向量化，在这里我们的类别只有两种，经过转化后标签数据变为二维
train_labels = np_utils.to_categorical(train_labels,nb_classes)
valid_labels = np_utils.to_categorical(valid_labels,nb_classes)
test_labels = np_utils.to_categorical(test_labels,nb_classes)
# 像素数据浮点化以便归一化
train_images = train_images.astype('float32')
valid_images = valid_images.astype('float32')
test_images = test_images.astype('float32')
#将其归一化,图像的各像素值归一化到0~1区间
train_images/=255
valid_images/=255
test_images/=255
#数据集先浮点后归一化的目的是提升网络收敛速度，减少训练时间，同时适应值域在（0,1）之间的激活函数，
# 增大区分度。其实归一化有一个特别重要的原因是确保特征值权重一致。举个例子，我们使用mse这样的均方误差函数时，
# 大的特征数值比如(5000-1000)2与小的特征值(3-1)2相加再求平均得到的误差值，显然大值对误差值的影响最大，但大部分情况下，
# 特征值的权重应该是一样的，只是因为单位不同才导致数值相差甚大。因此，我们提前对特征数据做归一化处理，以解决此类问题。
self.train_images = train_images
self.train_labels = train_labels
self.valid_images = valid_images
self.valid_labels = valid_labels
self.test_images = test_images
self.test_labels = test_labels
if __name__ == '__main__':
dataset = Dataset('D:/tmp/data/')
dataset.load()
model = Model()
model.build_model(dataset)
model.train(dataset)
model.save_model(file_path='me.face.model.h5')
加载数据，训练模型，保存模型，模型保存为model_weight.h5文件，Model类里的face_predict函数供外部接口调用，迎来识别人脸识别，返回类别。
3：外部接口调用。从视频里进行图像的采集，进而调用face_predict函数进行识别。用cv2的haarcascade_frontalface_alt2进行图像的采集。
#-*- coding: utf-8 -*-
import cv2
import sys
import gc
from face_train_use_keras import Model
from keras import backend as k
if __name__ == '__main__':
#加载模型
model = Model()
model.load_model(file_path = 'model_weight.h5')
#框住人脸的矩形边框颜色
color = (0, 255, 0)
#捕获指定摄像头的实时视频流
#cap = cv2.VideoCapture('D:/tmp/vedio/123.mp4')
cap = cv2.VideoCapture('D:/tmp/vedio/123.mp4')
#人脸识别分类器本地存储路径
cascade_path = "C:/projects/opencv-python/opencv/modules/objdetect/src/cascadedetect/haarcascade_frontalface_alt2.xml"
#循环检测识别人脸
while True:
ret, frame = cap.read() #读取一帧视频
#图像灰化，降低计算复杂度
frame_gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
#使用人脸识别分类器，读入分类器
cascade = cv2.CascadeClassifier(cascade_path)
#利用分类器识别出哪个区域为人脸
faceRects = cascade.detectMultiScale(frame_gray, scaleFactor = 1.2, minNeighbors = 3, minSize = (50, 50))
if len(faceRects) > 0:
i = 0
for faceRect in faceRects:
x, y, w, h = faceRect
#截取脸部图像提交给模型识别这是谁
image = frame[y - 10: y + h + 10, x - 10: x + w + 10]
import scipy.misc
faceID = model.face_predict(image)
import numpy as np
#如果是“我”
if faceID == 0:
cv2.rectangle(frame, (x - 10, y - 10), (x + w + 10, y + h + 10), color, thickness = 2)
#文字提示是谁
cv2.putText(frame,'mengmeng',
(x + 30, y + 30), #坐标
cv2.FONT_HERSHEY_SIMPLEX, #字体
1, #字号
(255,0,255), #颜色
2) #字的线宽
else:
cv2.rectangle(frame, (x - 10, y - 10), (x + w + 10, y + h + 10), color, thickness=2)
# 文字提示是谁
cv2.putText(frame, 'lailai',
(x + 30, y + 30), # 坐标
cv2.FONT_HERSHEY_SIMPLEX, # 字体
1, # 字号
(255, 0, 255), # 颜色
2)
cv2.imshow("识别朕", frame)
#等待10毫秒看是否有按键输入
k = cv2.waitKey(10)
#如果输入q则退出循环
if k & 0xFF == ord('q'):
break
#释放摄像头并销毁所有窗口
cap.release()
cv2.destroyAllWindows()
运行的结果如下：
首先是图像的采集效果。
然后通过采集的图像传给face_predict函数进行预测。绿色的框随着任务头像的移动而移动。进而显示预测的类别
识别的效果如上图。其实是一个完整的视频，会时时的识别。