|
|
# 实验环境：python 3.6 + opencv-python 3.4.14.51
|
|
|
|
|
|
import cv2
|
|
|
import numpy as np
|
|
|
import os
|
|
|
import shutil
|
|
|
import threading
|
|
|
import tkinter as tk
|
|
|
from PIL import Image, ImageTk
|
|
|
|
|
|
# 首先读取config文件，第一行代表当前已经储存的人名个数，接下来每一行是（id，name）标签和对应的人名
|
|
|
id_dict = {} # 字典里存的是id——name键值对
|
|
|
Total_face_num = 999 # 已经被识别有用户名的人脸个数,
|
|
|
|
|
|
|
|
|
def init(): # 将config文件内的信息读入到字典中
|
|
|
f = open('config.txt')
|
|
|
global Total_face_num
|
|
|
Total_face_num = int(f.readline())
|
|
|
|
|
|
for i in range(int(Total_face_num)):
|
|
|
line = f.readline()
|
|
|
id_name = line.split(' ')
|
|
|
id_dict[int(id_name[0])] = id_name[1]
|
|
|
f.close()
|
|
|
|
|
|
|
|
|
init()
|
|
|
|
|
|
# 加载OpenCV人脸检测分类器Haar
|
|
|
face_cascade = cv2.CascadeClassifier("haarcascade_frontalface_default.xml")
|
|
|
|
|
|
# 准备好识别方法LBPH方法
|
|
|
recognizer = cv2.face.LBPHFaceRecognizer_create()
|
|
|
|
|
|
# 打开标号为0的摄像头
|
|
|
camera = cv2.VideoCapture(0) # 摄像头
|
|
|
success, img = camera.read() # 从摄像头读取照片
|
|
|
W_size = 0.1 * camera.get(3)
|
|
|
H_size = 0.1 * camera.get(4)
|
|
|
|
|
|
system_state_lock = 0 # 标志系统状态的量 0表示无子线程在运行 1表示正在刷脸 2表示正在录入新面孔。
|
|
|
# 相当于mutex锁，用于线程同步
|
|
|
|
|
|
|
|
|
'''
|
|
|
============================================================================================
|
|
|
以上是初始化
|
|
|
============================================================================================
|
|
|
'''
|
|
|
|
|
|
|
|
|
def Get_new_face():
|
|
|
print("正在从摄像头录入新人脸信息 \n")
|
|
|
|
|
|
# 存在目录data就清空，不存在就创建，确保最后存在空的data目录
|
|
|
filepath = "data"
|
|
|
if not os.path.exists(filepath):
|
|
|
os.mkdir(filepath)
|
|
|
else:
|
|
|
shutil.rmtree(filepath)
|
|
|
os.mkdir(filepath)
|
|
|
|
|
|
sample_num = 0 # 已经获得的样本数
|
|
|
|
|
|
while True: # 从摄像头读取图片
|
|
|
|
|
|
global success
|
|
|
global img # 因为要显示在可视化的控件内，所以要用全局的
|
|
|
success, img = camera.read()
|
|
|
|
|
|
# 转为灰度图片
|
|
|
if success is True:
|
|
|
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
|
|
|
else:
|
|
|
break
|
|
|
|
|
|
# 检测人脸，将每一帧摄像头记录的数据带入OpenCv中，让Classifier判断人脸
|
|
|
# 其中gray为要检测的灰度图像，1.3为每次图像尺寸减小的比例，5为minNeighbors
|
|
|
faces = face_cascade.detectMultiScale(gray, 1.3, 5)
|
|
|
|
|
|
# 框选人脸，for循环保证一个能检测的实时动态视频流
|
|
|
for (x, y, w, h) in faces:
|
|
|
# xy为左上角的坐标,w为宽，h为高，用rectangle为人脸标记画框
|
|
|
cv2.rectangle(img, (x, y), (x + w, y + w), (255, 0, 0))
|
|
|
# 样本数加1
|
|
|
sample_num += 1
|
|
|
# 保存图像，把灰度图片看成二维数组来检测人脸区域，这里是保存在data缓冲文件夹内
|
|
|
T = Total_face_num
|
|
|
cv2.imwrite("./data/User." + str(T) + '.' + str(sample_num) + '.jpg', gray[y:y + h, x:x + w])
|
|
|
|
|
|
pictur_num = 1000 # 表示摄像头拍摄取样的数量,越多效果越好，但获取以及训练的越慢
|
|
|
|
|
|
cv2.waitKey(1)
|
|
|
if sample_num > pictur_num:
|
|
|
break
|
|
|
else: # 控制台内输出进度条
|
|
|
l = int(sample_num / pictur_num * 50)
|
|
|
r = int((pictur_num - sample_num) / pictur_num * 50)
|
|
|
print("\r" + "%{:.1f}".format(sample_num / pictur_num * 100) + "=" * l + "->" + "_" * r, end="")
|
|
|
var.set("%{:.1f}".format(sample_num / pictur_num * 100)) # 控件可视化进度信息
|
|
|
# tk.Tk().update()
|
|
|
window.update() # 刷新控件以实时显示进度
|
|
|
|
|
|
|
|
|
def Train_new_face():
|
|
|
print("\n正在训练")
|
|
|
# cv2.destroyAllWindows()
|
|
|
path = 'data'
|
|
|
|
|
|
# 初始化识别的方法
|
|
|
recog = recognizer
|
|
|
|
|
|
# 调用函数并将数据喂给识别器训练
|
|
|
faces, ids = get_images_and_labels(path)
|
|
|
print('本次用于训练的识别码为:') # 调试信息
|
|
|
print(ids) # 输出识别码
|
|
|
|
|
|
# 训练模型 #将输入的所有图片转成四维数组
|
|
|
recog.train(faces, np.array(ids))
|
|
|
# 保存模型
|
|
|
|
|
|
yml = str(Total_face_num) + ".yml"
|
|
|
rec_f = open(yml, "w+")
|
|
|
rec_f.close()
|
|
|
recog.save(yml)
|
|
|
|
|
|
# recog.save('aaa.yml')
|
|
|
|
|
|
|
|
|
# 创建一个函数，用于从数据集文件夹中获取训练图片,并获取id
|
|
|
# 注意图片的命名格式为User.id.sampleNum
|
|
|
def get_images_and_labels(path):
|
|
|
image_paths = [os.path.join(path, f) for f in os.listdir(path)]
|
|
|
# 新建连个list用于存放
|
|
|
face_samples = []
|
|
|
ids = []
|
|
|
|
|
|
# 遍历图片路径，导入图片和id添加到list中
|
|
|
for image_path in image_paths:
|
|
|
|
|
|
# 通过图片路径将其转换为灰度图片
|
|
|
img = Image.open(image_path).convert('L')
|
|
|
|
|
|
# 将图片转化为数组
|
|
|
img_np = np.array(img, 'uint8')
|
|
|
|
|
|
if os.path.split(image_path)[-1].split(".")[-1] != 'jpg':
|
|
|
continue
|
|
|
|
|
|
# 为了获取id，将图片和路径分裂并获取
|
|
|
id = int(os.path.split(image_path)[-1].split(".")[1])
|
|
|
|
|
|
# 调用熟悉的人脸分类器
|
|
|
detector = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
|
|
|
|
|
|
faces = detector.detectMultiScale(img_np)
|
|
|
|
|
|
# 将获取的图片和id添加到list中
|
|
|
for (x, y, w, h) in faces:
|
|
|
face_samples.append(img_np[y:y + h, x:x + w])
|
|
|
ids.append(id)
|
|
|
return face_samples, ids
|
|
|
|
|
|
|
|
|
def write_config():
|
|
|
print("新人脸训练结束")
|
|
|
f = open('config.txt', "a")
|
|
|
T = Total_face_num
|
|
|
f.write(str(T) + " User" + str(T) + " \n")
|
|
|
f.close()
|
|
|
id_dict[T] = "User" + str(T)
|
|
|
|
|
|
# 这里修改文件的方式是先读入内存，然后修改内存中的数据，最后写回文件
|
|
|
f = open('config.txt', 'r+')
|
|
|
flist = f.readlines()
|
|
|
flist[0] = str(int(flist[0]) + 1) + " \n"
|
|
|
f.close()
|
|
|
|
|
|
f = open('config.txt', 'w+')
|
|
|
f.writelines(flist)
|
|
|
f.close()
|
|
|
|
|
|
|
|
|
'''
|
|
|
============================================================================================
|
|
|
以上是录入新人脸信息功能的实现
|
|
|
============================================================================================
|
|
|
'''
|
|
|
|
|
|
|
|
|
def scan_face():
|
|
|
# 使用之前训练好的模型
|
|
|
for i in range(Total_face_num): # 每个识别器都要用
|
|
|
i += 1
|
|
|
yml = str(i) + ".yml"
|
|
|
print("\n本次:" + yml) # 调试信息
|
|
|
recognizer.read(yml)
|
|
|
|
|
|
ave_poss = 0
|
|
|
for times in range(10): # 每个识别器扫描十遍
|
|
|
times += 1
|
|
|
cur_poss = 0
|
|
|
global success
|
|
|
global img
|
|
|
|
|
|
global system_state_lock
|
|
|
while system_state_lock == 2: # 如果正在录入新面孔就阻塞
|
|
|
print("\r刷脸被录入面容阻塞", end="")
|
|
|
pass
|
|
|
|
|
|
success, img = camera.read()
|
|
|
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
|
|
|
# 识别人脸
|
|
|
faces = face_cascade.detectMultiScale(
|
|
|
gray,
|
|
|
scaleFactor=1.2,
|
|
|
minNeighbors=5,
|
|
|
minSize=(int(W_size), int(H_size))
|
|
|
)
|
|
|
# 进行校验
|
|
|
for (x, y, w, h) in faces:
|
|
|
|
|
|
# global system_state_lock
|
|
|
while system_state_lock == 2: # 如果正在录入新面孔就阻塞
|
|
|
print("\r刷脸被录入面容阻塞", end="")
|
|
|
pass
|
|
|
# 这里调用Cv2中的rectangle函数 在人脸周围画一个矩形
|
|
|
cv2.rectangle(img, (x, y), (x + w, y + h), (0, 255, 0), 2)
|
|
|
# 调用分类器的预测函数，接收返回值标签和置信度
|
|
|
idnum, confidence = recognizer.predict(gray[y:y + h, x:x + w])
|
|
|
conf = confidence
|
|
|
# 计算出一个检验结果
|
|
|
if confidence < 100: # 可以识别出已经训练的对象——直接输出姓名在屏幕上
|
|
|
if idnum in id_dict:
|
|
|
user_name = id_dict[idnum]
|
|
|
else:
|
|
|
# print("无法识别的ID:{}\t".format(idnum), end="")
|
|
|
user_name = "Untagged user:" + str(idnum)
|
|
|
confidence = round(100 - confidence)
|
|
|
else: # 无法识别此对象，那么就开始训练
|
|
|
user_name = "unknown"
|
|
|
# print("检测到陌生人脸\n")
|
|
|
|
|
|
# cv2.destroyAllWindows()
|
|
|
# global Total_face_num
|
|
|
# Total_face_num += 1
|
|
|
# Get_new_face() # 采集新人脸
|
|
|
# Train_new_face() # 训练采集到的新人脸
|
|
|
# write_config() # 修改配置文件
|
|
|
# recognizer.read('aaa.yml') # 读取新识别器
|
|
|
|
|
|
# 加载一个字体用于输出识别对象的信息
|
|
|
font = cv2.FONT_HERSHEY_SIMPLEX
|
|
|
|
|
|
# 输出检验结果以及用户名
|
|
|
cv2.putText(img, str(user_name), (x + 5, y - 5), font, 1, (0, 0, 255), 1)
|
|
|
cv2.putText(img, str(confidence), (x + 5, y + h - 5), font, 1, (0, 0, 0), 1)
|
|
|
|
|
|
# 展示结果
|
|
|
# cv2.imshow('camera', img)
|
|
|
|
|
|
print("conf=" + str(conf), end="\t")
|
|
|
if 15 > conf > 0:
|
|
|
cur_poss = 1 # 表示可以识别
|
|
|
elif 60 > conf > 35:
|
|
|
cur_poss = 1 # 表示可以识别
|
|
|
else:
|
|
|
cur_poss = 0 # 表示不可以识别
|
|
|
|
|
|
k = cv2.waitKey(1)
|
|
|
if k == 27:
|
|
|
# cam.release() # 释放资源
|
|
|
cv2.destroyAllWindows()
|
|
|
break
|
|
|
|
|
|
ave_poss += cur_poss
|
|
|
|
|
|
if ave_poss >= 5: # 有一半以上识别说明可行则返回
|
|
|
return i
|
|
|
|
|
|
return 0 # 全部过一遍还没识别出说明无法识别
|
|
|
|
|
|
|
|
|
'''
|
|
|
============================================================================================
|
|
|
以上是关于刷脸功能的设计
|
|
|
============================================================================================
|
|
|
'''
|
|
|
|
|
|
|
|
|
def f_scan_face_thread():
|
|
|
# 使用之前训练好的模型
|
|
|
# recognizer.read('aaa.yml')
|
|
|
var.set('刷脸')
|
|
|
ans = scan_face()
|
|
|
if ans == 0:
|
|
|
print("最终结果：无法识别")
|
|
|
var.set("最终结果：无法识别")
|
|
|
|
|
|
else:
|
|
|
ans_name = "最终结果：" + str(ans) + id_dict[ans]
|
|
|
print(ans_name)
|
|
|
var.set(ans_name)
|
|
|
|
|
|
global system_state_lock
|
|
|
print("锁被释放0")
|
|
|
system_state_lock = 0 # 修改system_state_lock,释放资源
|
|
|
|
|
|
|
|
|
def f_scan_face():
|
|
|
global system_state_lock
|
|
|
print("\n当前锁的值为：" + str(system_state_lock))
|
|
|
if system_state_lock == 1:
|
|
|
print("阻塞，因为正在刷脸")
|
|
|
return 0
|
|
|
elif system_state_lock == 2: # 如果正在录入新面孔就阻塞
|
|
|
print("\n刷脸被录入面容阻塞\n"
|
|
|
"")
|
|
|
return 0
|
|
|
system_state_lock = 1
|
|
|
p = threading.Thread(target=f_scan_face_thread)
|
|
|
p.setDaemon(True) # 把线程P设置为守护线程 若主线程退出 P也跟着退出
|
|
|
p.start()
|
|
|
|
|
|
|
|
|
def f_rec_face_thread():
|
|
|
var.set('录入')
|
|
|
cv2.destroyAllWindows()
|
|
|
global Total_face_num
|
|
|
Total_face_num += 1
|
|
|
Get_new_face() # 采集新人脸
|
|
|
print("采集完毕，开始训练")
|
|
|
global system_state_lock # 采集完就可以解开锁
|
|
|
print("锁被释放0")
|
|
|
system_state_lock = 0
|
|
|
|
|
|
Train_new_face() # 训练采集到的新人脸
|
|
|
write_config() # 修改配置文件
|
|
|
|
|
|
|
|
|
# recognizer.read('aaa.yml') # 读取新识别器
|
|
|
|
|
|
# global system_state_lock
|
|
|
# print("锁被释放0")
|
|
|
# system_state_lock = 0 # 修改system_state_lock,释放资源
|
|
|
|
|
|
|
|
|
def f_rec_face():
|
|
|
global system_state_lock
|
|
|
print("当前锁的值为：" + str(system_state_lock))
|
|
|
if system_state_lock == 2:
|
|
|
print("阻塞，因为正在录入面容")
|
|
|
return 0
|
|
|
else:
|
|
|
system_state_lock = 2 # 修改system_state_lock
|
|
|
print("改为2", end="")
|
|
|
print("当前锁的值为：" + str(system_state_lock))
|
|
|
|
|
|
p = threading.Thread(target=f_rec_face_thread)
|
|
|
p.setDaemon(True) # 把线程P设置为守护线程 若主线程退出 P也跟着退出
|
|
|
p.start()
|
|
|
# tk.Tk().update()
|
|
|
|
|
|
|
|
|
# system_state_lock = 0 # 修改system_state_lock,释放资源
|
|
|
|
|
|
|
|
|
def f_exit(): # 退出按钮
|
|
|
exit()
|
|
|
|
|
|
|
|
|
'''
|
|
|
============================================================================================
|
|
|
以上是关于多线程的设计
|
|
|
============================================================================================
|
|
|
'''
|
|
|
|
|
|
window = tk.Tk()
|
|
|
window.title('xichengzhi\' Face_rec 3.0') # 窗口标题
|
|
|
window.geometry('1000x500') # 这里的乘是小x
|
|
|
|
|
|
# 在图形界面上设定标签，类似于一个提示窗口的作用
|
|
|
var = tk.StringVar()
|
|
|
l = tk.Label(window, textvariable=var, bg='green', fg='white', font=('Arial', 12), width=50, height=4)
|
|
|
# 说明： bg为背景，fg为字体颜色，font为字体，width为长，height为高，这里的长和高是字符的长和高，比如height=2,就是标签有2个字符这么高
|
|
|
l.pack() # 放置l控件
|
|
|
var.set('人脸识别')
|
|
|
|
|
|
# 在窗口界面设置放置Button按键并绑定处理函数
|
|
|
button_a = tk.Button(window, text='开始刷脸', font=('Arial', 12), width=10, height=2, command=f_scan_face)
|
|
|
button_a.place(x=800, y=120)
|
|
|
|
|
|
button_b = tk.Button(window, text='录入人脸', font=('Arial', 12), width=10, height=2, command=f_rec_face)
|
|
|
button_b.place(x=800, y=220)
|
|
|
|
|
|
button_c = tk.Button(window, text='退出', font=('Arial', 12), width=10, height=2, command=f_exit)
|
|
|
button_c.place(x=800, y=320)
|
|
|
|
|
|
panel = tk.Label(window, width=500, height=350) # 摄像头模块大小
|
|
|
panel.place(x=10, y=100) # 摄像头模块的位置
|
|
|
window.config(cursor="arrow")
|
|
|
|
|
|
|
|
|
def video_loop(): # 用于在label内动态展示摄像头内容（摄像头嵌入控件）
|
|
|
# success, img = camera.read() # 从摄像头读取照片
|
|
|
global success
|
|
|
global img
|
|
|
if success:
|
|
|
cv2.waitKey(1)
|
|
|
cv2image = cv2.cvtColor(img, cv2.COLOR_BGR2RGBA) # 转换颜色从BGR到RGBA
|
|
|
current_image = Image.fromarray(cv2image) # 将图像转换成Image对象
|
|
|
imgtk = ImageTk.PhotoImage(image=current_image)
|
|
|
panel.imgtk = imgtk
|
|
|
panel.config(image=imgtk)
|
|
|
window.after(1, video_loop)
|
|
|
|
|
|
|
|
|
video_loop()
|
|
|
|
|
|
# 窗口循环，用于显示
|
|
|
window.mainloop()
|
|
|
|
|
|
'''
|
|
|
============================================================================================
|
|
|
以上是关于界面的设计
|
|
|
============================================================================================
|
|
|
'''
|
