第一句子网 > Python+dlib+opencv实现简单的人脸识别

Python+dlib+opencv实现简单的人脸识别

时间：2018-12-25 15:36:36

一、建立人脸数据集

1.采集人脸

建立自己的人脸数据集：建议采集多角度的20张人脸

import cv2import dlibimport osimport sysimport random# 存储位置output_dir = 'D:/face/631907060410/4me' #采集人脸的路径：其中2***为人名（英文方式）方便对比，第二次编译时修改为3***，进行多人比对size = 100if not os.path.exists(output_dir):os.makedirs(output_dir)# 改变图片的亮度与对比度def relight(img, light=1, bias=0):w = img.shape[1]h = img.shape[0]#image = []for i in range(0,w):for j in range(0,h):for c in range(3):tmp = int(img[j,i,c]*light + bias)if tmp > 255:tmp = 255elif tmp < 0:tmp = 0img[j,i,c] = tmpreturn img#使用dlib自带的frontal_face_detector作为我们的特征提取器detector = dlib.get_frontal_face_detector()# 打开摄像头参数为输入流，可以为摄像头或视频文件camera = cv2.VideoCapture(0) #打开摄像头方式#camera = cv2.VideoCapture('2.mp4') #采集视频的人脸index = 1while True:if (index <= 20):#存储20张人脸特征图像print('Being processed picture %s' % index)# 从摄像头读取照片success, img = camera.read()# 转为灰度图片gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)# 使用detector进行人脸检测dets = detector(gray_img, 1)for i, d in enumerate(dets):x1 = d.top() if d.top() > 0 else 0y1 = d.bottom() if d.bottom() > 0 else 0x2 = d.left() if d.left() > 0 else 0y2 = d.right() if d.right() > 0 else 0face = img[x1:y1,x2:y2]# 调整图片的对比度与亮度，对比度与亮度值都取随机数，这样能增加样本的多样性face = relight(face, random.uniform(0.5, 1.5), random.randint(-50, 50))face = cv2.resize(face, (size,size))cv2.imshow('image', face)cv2.imwrite(output_dir+'/'+str(index)+'.jpg', face)index += 1key = cv2.waitKey(30) & 0xffif key == 27:breakelse:print('Finished!')# 释放摄像头 release cameracamera.release()# 删除建立的窗口 delete all the windowscv2.destroyAllWindows()break

我采集了三组：

2.采集对应20张图片的68个特征点数组和平均特征数组

由于我采集了三组数据，所以会有60个数组（当然，如果有无效图的话就没有60组了）。

当光线过亮或者过暗，五官没有在采集到的图片内或模糊不清时，这张图就无效

from cv2 import cv2 as cv2import osimport dlibfrom skimage import ioimport csvimport numpy as np# 要读取人脸图像文件的路径path_images_from_camera = "D:/face/631907060410/"# Dlib 正向人脸检测器detector = dlib.get_frontal_face_detector()# Dlib 人脸预测器predictor = dlib.shape_predictor("C:/Users/86150/JupyterProject/shape_predictor_68_face_landmarks.dat") #我这是在同一路径下的，shape_predictor_68_face_landmarks.dat不在同一路径的话#要写为绝对路径："D:/****/****/shape_predictor_68_face_landmarks.dat"# Dlib 人脸识别模型# Face recognition model, the object maps human faces into 128D vectorsface_rec = dlib.face_recognition_model_v1("C:/Users/86150/JupyterProject/dlib_face_recognition_resnet_model_v1.dat")#我这是在同一路径下的，dlib_face_recognition_resnet_model_v1.dat不在#要写为绝对路径："D:/****/****/dlib_face_recognition_resnet_model_v1.dat"# 返回单张图像的 128D 特征def return_128d_features(path_img):img_rd = io.imread(path_img)s=path_imga=s[21:22]i1=str(a)a1=s[24:]str1="/"b=a1[a1.index(str1):-4]b1=b[1:]i2=str(b1)img_gray = cv2.cvtColor(img_rd, cv2.COLOR_BGR2RGB)faces = detector(img_gray, 1)for i in range(len(faces)):landmarks = np.matrix([[p.x, p.y] for p in predictor(img_rd,faces[i]).parts()]) for idx, point in enumerate(landmarks):# 68点的坐标pos = (point[0, 0], point[0, 1])add="D:/face/631907060410/face"+i1+"_feature"+i2+".csv"with open(add, "a", newline="") as csvfile:writer1 = csv.writer(csvfile)writer1.writerow((idx,pos))print(add)print("%-40s %-20s" % ("检测到人脸的图像 / image with faces detected:", path_img), '\n')# 因为有可能截下来的人脸再去检测，检测不出来人脸了# 所以要确保是检测到人脸的人脸图像拿去算特征if len(faces) != 0:shape = predictor(img_gray, faces[0])face_descriptor = pute_face_descriptor(img_gray, shape)else:face_descriptor = 0print("no face")return face_descriptor# 将文件夹中照片特征提取出来, 写入 CSVdef return_features_mean_personX(path_faces_personX):features_list_personX = []s=path_faces_personXa=s[21:22]i2=str(a)photos_list = os.listdir(path_faces_personX)if photos_list:for i in range(len(photos_list)):# 调用return_128d_features()得到128d特征print("%-40s %-20s" % ("正在读的人脸图像 / image to read:", path_faces_personX + "/" + photos_list[i]))features_128d = return_128d_features(path_faces_personX + "/" + photos_list[i])# print(features_128d)# 遇到没有检测出人脸的图片跳过if features_128d == 0:i += 1else:features_list_personX.append(features_128d)i1=str(i+1)add="D:/face/feature/"+i2+"feature"+i1+".csv"print(add)with open(add, "w", newline="") as csvfile:writer1 = csv.writer(csvfile)writer1.writerow(features_128d)else:print("文件夹内图像文件为空 / Warning: No images in " + path_faces_personX + '/', '\n')# 计算 128D 特征的均值# N x 128D -> 1 x 128Dif features_list_personX:features_mean_personX = np.array(features_list_personX).mean(axis=0)else:features_mean_personX = '0'return features_mean_personX# 读取某人所有的人脸图像的数据people = os.listdir(path_images_from_camera)people.sort()with open("D:/face/631907060410/face_feature_mean.csv", "w", newline="") as csvfile: #程序会新建一个表格文件来保存特征值，方便以后比对writer = csv.writer(csvfile)for person in people:print("##### " + person + " #####")# Get the mean/average features of face/personX, it will be a list with a length of 128Dfeatures_mean_personX = return_features_mean_personX(path_images_from_camera + person)writer.writerow(features_mean_personX)print("特征均值 / The mean of features:", list(features_mean_personX))print('\n')print("所有录入人脸数据存入 / Save all the features of faces registered into: D:/631907060410face_feature_mean.csv")

生成的60组特征点数组和平均特征数组

特征点数组（以face2_feature1.csv为例）：

每张图片的特征值（2feature1.csv为例）：

平均特征值：

二、人脸识别

import osimport winsound # 系统音效from playsound import playsound # 音频播放import dlib# 人脸处理的库 Dlibimport csv # 存入表格import timeimport sysimport numpy as np # 数据处理的库 numpyfrom cv2 import cv2 as cv2 # 图像处理的库 OpenCvimport pandas as pd # 数据处理的库 Pandas# 人脸识别模型，提取128D的特征矢量# face recognition model, the object maps human faces into 128D vectors# Refer this tutorial: /python/index.html#dlib.face_recognition_model_v1facerec = dlib.face_recognition_model_v1("C:/Users/86150/JupyterProject/dlib_face_recognition_resnet_model_v1.dat")#我这是在同一路径下的，dlib_face_recognition_resnet_model_v1.dat不在#要写为绝对路径："D:/****/****/dlib_face_recognition_resnet_model_v1.dat"# 计算两个128D向量间的欧式距离# compute the e-distance between two 128D featuresdef return_euclidean_distance(feature_1, feature_2):feature_1 = np.array(feature_1)feature_2 = np.array(feature_2)dist = np.sqrt(np.sum(np.square(feature_1 - feature_2)))return dist# 处理存放所有人脸特征的 csvpath_features_known_csv = "D:/face/631907060410/face_feature_mean.csv"csv_rd = pd.read_csv(path_features_known_csv, header=None)# 用来存放所有录入人脸特征的数组# the array to save the features of faces in the databasefeatures_known_arr = []# 读取已知人脸数据# print known facesfor i in range(csv_rd.shape[0]):features_someone_arr = []for j in range(0, len(csv_rd.ix[i, :])):features_someone_arr.append(csv_rd.ix[i, :][j])features_known_arr.append(features_someone_arr)print("Faces in Database：", len(features_known_arr))# Dlib 检测器和预测器# The detector and predictor will be useddetector = dlib.get_frontal_face_detector()predictor = dlib.shape_predictor('C:/Users/86150/JupyterProject/shape_predictor_68_face_landmarks.dat')#我这是在同一路径下的，shape_predictor_68_face_landmarks.dat不在#要写为绝对路径："D:/****/****/shape_predictor_68_face_landmarks.dat# 创建 cv2 摄像头对象# cv2.VideoCapture(0) to use the default camera of PC,# and you can use local video name by use cv2.VideoCapture(filename)cap = cv2.VideoCapture(0)# cap.set(propId, value)# 设置视频参数，propId 设置的视频参数，value 设置的参数值cap.set(3, 480)# cap.isOpened() 返回 true/false 检查初始化是否成功# when the camera is openwhile cap.isOpened():flag, img_rd = cap.read()kk = cv2.waitKey(1)# 取灰度img_gray = cv2.cvtColor(img_rd, cv2.COLOR_RGB2GRAY)# 人脸数 facesfaces = detector(img_gray, 0)# 待会要写的字体 font to write laterfont = cv2.FONT_HERSHEY_COMPLEX# 存储当前摄像头中捕获到的所有人脸的坐标/名字# the list to save the positions and names of current faces capturedpos_namelist = []name_namelist = []# 按下 q 键退出# press 'q' to exitif kk == ord('q'):breakelse:# 检测到人脸 when face detectedif len(faces) != 0: # 获取当前捕获到的图像的所有人脸的特征，存储到 features_cap_arr# get the features captured and save into features_cap_arrfeatures_cap_arr = []for i in range(len(faces)):shape = predictor(img_rd, faces[i])features_cap_arr.append(pute_face_descriptor(img_rd, shape))# 遍历捕获到的图像中所有的人脸# traversal all the faces in the databasefor k in range(len(faces)):print("##### camera person", k+1, "#####")# 让人名跟随在矩形框的下方# 确定人名的位置坐标# 先默认所有人不认识，是 unknown# set the default names of faces with "unknown"name_namelist.append("unknown")# 每个捕获人脸的名字坐标 the positions of faces capturedpos_namelist.append(tuple([faces[k].left(), int(faces[k].bottom() + (faces[k].bottom() - faces[k].top())/4)]))# 对于某张人脸，遍历所有存储的人脸特征# for every faces detected, compare the faces in the databasee_distance_list = []for i in range(len(features_known_arr)):# 如果 person_X 数据不为空if str(features_known_arr[i][0]) != '0.0':print("with person", str(i + 1), "the e distance: ", end='')e_distance_tmp = return_euclidean_distance(features_cap_arr[k], features_known_arr[i])print(e_distance_tmp)e_distance_list.append(e_distance_tmp)else:# 空数据 person_Xe_distance_list.append(999999999)# 找出最接近的一个人脸数据是第几个# Find the one with minimum e distancesimilar_person_num = e_distance_list.index(min(e_distance_list))print("Minimum e distance with person", int(similar_person_num)+1)# 计算人脸识别特征与数据集特征的欧氏距离# 距离小于0.4则标出为可识别人物if min(e_distance_list) < 0.4:# 这里可以修改摄像头中标出的人名# Here you can modify the names shown on the camera# 1、遍历文件夹目录folder_name = 'D:/face/631907060410'# 最接近的人脸sum=similar_person_num+1key_id=1 # 从第一个人脸数据文件夹进行对比# 获取文件夹中的文件名:1wang、2zhou、3...file_names = os.listdir(folder_name)for name in file_names:# print(name+'->'+str(key_id))if sum ==key_id:#winsound.Beep(300,500)# 响铃：300频率，500持续时间name_namelist[k] = name[1:]#人名删去第一个数字（用于视频输出标识）key_id += 1# 播放欢迎光临音效#playsound('D:/myworkspace/JupyterNotebook/People/music/welcome.wav')# print("May be person "+str(int(similar_person_num)+1))# -----------筛选出人脸并保存到visitor文件夹------------for i, d in enumerate(faces):x1 = d.top() if d.top() > 0 else 0y1 = d.bottom() if d.bottom() > 0 else 0x2 = d.left() if d.left() > 0 else 0y2 = d.right() if d.right() > 0 else 0face = img_rd[x1:y1,x2:y2]size = 64face = cv2.resize(face, (size,size))# 要存储visitor人脸图像文件的路径path_visitors_save_dir = "D:/face/known" #自己在faces路径下先建一个known文件，否则可能会报错# 存储格式：-06-24-14-33-40wang.jpgnow_time = time.strftime("%Y-%m-%d-%H-%M-%S", time.localtime())save_name = str(now_time)+str(name_namelist[k])+'.jpg'# print(save_name)# 本次图片保存的完整urlsave_path = path_visitors_save_dir+'/'+ save_name # 遍历visitor文件夹所有文件名visitor_names = os.listdir(path_visitors_save_dir)visitor_name=''for name in visitor_names:# 名字切片到分钟数：-06-26-11-33-00wangyu.jpgvisitor_name=(name[0:16]+'-00'+name[19:])# print(visitor_name)visitor_save=(save_name[0:16]+'-00'+save_name[19:])# print(visitor_save)# 一分钟之内重复的人名不保存if visitor_save!=visitor_name:cv2.imwrite(save_path, face)print('新存储：'+path_visitors_save_dir+'/'+str(now_time)+str(name_namelist[k])+'.jpg')else:print('重复，未保存！')else:# 播放无法识别音效#playsound('D:/myworkspace/JupyterNotebook/People/music/sorry.wav')print("Unknown person")# -----保存图片-------# -----------筛选出人脸并保存到visitor文件夹------------for i, d in enumerate(faces):x1 = d.top() if d.top() > 0 else 0y1 = d.bottom() if d.bottom() > 0 else 0x2 = d.left() if d.left() > 0 else 0y2 = d.right() if d.right() > 0 else 0face = img_rd[x1:y1,x2:y2]size = 64face = cv2.resize(face, (size,size))# 要存储visitor-》unknown人脸图像文件的路径path_visitors_save_dir = "D:/face/unknown"#自己在faces路径下先建一个unknown文件，否则可能会报错# 存储格式：-06-24-14-33-40unknown.jpgnow_time = time.strftime("%Y-%m-%d-%H-%M-%S", time.localtime())# print(save_name)# 本次图片保存的完整urlsave_path = path_visitors_save_dir+'/'+ str(now_time)+'unknown.jpg'cv2.imwrite(save_path, face)print('新存储：'+path_visitors_save_dir+'/'+str(now_time)+'unknown.jpg')# 矩形框# draw rectanglefor kk, d in enumerate(faces):# 绘制矩形框cv2.rectangle(img_rd, tuple([d.left(), d.top()]), tuple([d.right(), d.bottom()]), (0, 255, 255), 2)print('\n')# 在人脸框下面写人脸名字# write names under rectanglefor i in range(len(faces)):cv2.putText(img_rd, name_namelist[i], pos_namelist[i], font, 0.8, (0, 255, 255), 1, cv2.LINE_AA)print("Faces in camera now:", name_namelist, "\n")#cv2.putText(img_rd, "Press 'q': Quit", (20, 450), font, 0.8, (84, 255, 159), 1, cv2.LINE_AA)cv2.putText(img_rd, "Face Recognition", (20, 40), font, 1, (0, 0, 255), 1, cv2.LINE_AA)cv2.putText(img_rd, "Visitors: " + str(len(faces)), (20, 100), font, 1, (0, 0, 255), 1, cv2.LINE_AA)# 窗口显示 show with opencvcv2.imshow("camera", img_rd)k = cv2.waitKey(1)if k == 27: # press 'ESC' to quitbreak# 释放摄像头 release cameracap.release()# 删除建立的窗口 delete all the windowscv2.destroyAllWindows()

人脸识别成功：

选了两张采集时没有采集到文件的照片测试

成功识别。

再用其他人（没采集）的照片试试：

因为没有采集他的人脸特征，所以不能识别。

打开known和unknown文件夹看一下：