手把手教你实现实时人脸识别，带你刷脸刷颜值

dlib库的安装

问题

dlib原本是一个开源的c++人脸识别库，在github上接近万星，是一个非常优秀的开源框架。对于dlib库的c++版安装是比较麻烦的，python版dlib库安装网上资料一大堆，可以这样说，百分之九十九都是错的（不要问我为什么，我在本机和服务器上亲自试验过几遍）。下面，我将说明dlib库的安装流程。

解决方案

要知道一点，搜索引擎首选google（百度是真的烂），至于google的问题，还是百度一下“谷歌访问助手”，下载一下这个插件，然后你就可以免费访问google了（这是访问google最简单的方式了）。

我这里要说的是，直接看dlib的官方文档。无论谁写的安装教程，都不如官方文档来的权威，并且大大节省了你在搜索引擎上花的时间。

直接看图，官网说的是直接pip安装就ok了，但是麻烦的地方是需要c++的环境，还有cmake。所以，看下面的安装步骤

1、保证你的windows有c++的环境，建议你安装visual studio 2017，具体安装流程看这里https://www.jianshu.com/p/ed645d7b2c8a

2、安装一个anaconda，我建议大家都安装这个神器，他自带的虚拟环境管理工具是非常棒的，并且现在的人工智能环境下，基本都是用他，不多说，谁用谁知道。不过conda官网下载会非常慢，建议镜像下载。

3、接下来是安装cmake，很简单 pip install cmake就搞定了

4、最后就是安装dlib了，直接pip install dlib。这一步很可能会有一个大坑，你安装是能安装上的，但是多半import dlib会出错，是因为dlib版本太高了，pip默认下载的最新版dlib为19.18.0，我们应该换成19.17.0，所以，安装命令为：pip install dlib==19.17.0 别看这一个小小的错误，把我们组的一个技术老大都搞了一天。我算幸运，我安装的时候还没有19.18.0这个版本。

上面就是dlib安装的所有步骤了，你搜索很多博客会发现，让你装各种东西，什么boost这些，根本就不需要，官网都没有。

开始人脸验证

问题

人脸验证的问题，实际上就是我们过车站安检的过程，拿着你的身份证，摆个pose，看看身份证上的照片是不是您，也就是看你是否照骗。江湖也称这个过程为1：1人脸验证。

可别看这个小小的过程，实际包含了很多的细节，在照片的预处理中，就有几个环节。

1、人脸截取

在一张你的照片中，定位到你的人脸，然后截取人脸。
2、人脸校准
你拍的照片可能脸是侧脸，头是扬起来的，这样会影响模型的识别准确率，所以需要把人脸摆正

这里我结合opencv，写了一个实时人脸定位+校准的代码

import cv2
import dlib
# from skimage import io
import time
​
vc = cv2.VideoCapture(0)
# 使用特征提取器get_frontal_face_detector
detector = dlib.get_frontal_face_detector()
# dlib的68点模型，使用作者训练好的特征预测器
predictor = dlib.shape_predictor("shape_predictor_68_face_landmarks.dat")
​
​
while True:
    # Grab a single frame of video
    ret, img = vc.read()
    # img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    # 特征提取器的实例化
​
    start = time.time()
    # dets = detector(img, 1)
    # print(dets)
    # print(time.time()-start)
    start = time.time()
    dets, scores, idx = detector.run(img, 1, 0)
    print(type(scores))
    print(dets, scores)
    print(time.time()-start)
    # print(dets)
    # print("人脸数：", len(dets))
​
    for k, d in enumerate(dets):
            # print("第", k+1, "个人脸d的坐标：",
            #       "left:", d.left(),
            #       "right:", d.right(),
            #       "top:", d.top(),
            #       "bottom:", d.bottom())
​
            width = d.right() - d.left()
            heigth = d.bottom() - d.top()
​
            # print('人脸面积为：',(width*heigth))
            # 利用预测器预测
            shape = predictor(img, d)
            # 标出68个点的位置
            for i in range(68):
                cv2.circle(img, (shape.part(i).x, shape.part(i).y), 4, (0, 255, 0), -1, 8)
                cv2.putText(img, "", (shape.part(i).x, shape.part(i).y), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 255, 255))
            # 显示一下处理的图片，然后销毁窗口
            # Display the resulting image
            cv2.imshow('Video', img)
​
            # Hit 'q' on the keyboard to quit!
            if cv2.waitKey(1) & 0xFF == ord('q'):
                break
# Release handle to the webcam
video_capture.release()
cv2.destroyAllWindows()

其中有个模型文件需要去dlib官网下载http://dlib.net/files/
只要下载了模型，把代码拷贝到你的IDE，就能直接运行了

人脸验证

把你自己的模板图片，再给自己照一张图片，输入模型，就能进行人脸验证了，结果返回的是2个128维的向量，计算他们的欧式距离，就能算出相似度了。dlib库用的模型是34层的resnet网络，我自己测试过多次，效果不错。不过要想达到工程能用的水平，自己得去调优。

这里我也写了一个实时得人脸验证代码，老规矩，直接拷贝就能运行

import face_recognition
import cv2
import numpy as np
# This is a demo of running face recognition on live video from your webcam. It's a little more complicated than the
# other example, but it includes some basic performance tweaks to make things run a lot faster:
#   1. Process each video frame at 1/4 resolution (though still display it at full resolution)
#   2. Only detect faces in every other frame of video.
# PLEASE NOTE: This example requires OpenCV (the `cv2` library) to be installed only to read from your webcam.
# OpenCV is *not* required to use the face_recognition library. It's only required if you want to run this
# specific demo. If you have trouble installing it, try any of the other demos that don't require it instead.
# Get a reference to webcam #0 (the default one)
video_capture = cv2.VideoCapture(0)
# Load a sample picture and learn how to recognize it.
obama_image = face_recognition.load_image_file("cut.jpg")
# print(face_recognition.face_encodings(obama_image))
face_locations = face_recognition.face_locations(obama_image)
# print(face_locations)
obama_face_encoding = face_recognition.face_encodings(obama_image, face_locations)[0]
# Load a second sample picture and learn how to recognize it.
# biden_image = face_recognition.load_image_file("biden.jpg")
# biden_face_encoding = face_recognition.face_encodings(biden_image)[0]
# Create arrays of known face encodings and their names
known_face_encodings = [
    obama_face_encoding
    # biden_face_encoding
]
known_face_names = [
    "zh"
]
# Initialize some variables
face_locations = []
face_encodings = []
face_names = []
process_this_frame = True
​
while True:
    # Grab a single frame of video
    ret, frame = video_capture.read()
    # Resize frame of video to 1/4 size for faster face recognition processing
    small_frame = cv2.resize(frame, (0, 0), fx=1, fy=1)
    # Convert the image from BGR color (which OpenCV uses) to RGB color (which face_recognition uses)
    rgb_small_frame = small_frame[:, :, ::-1]
    # Only process every other frame of video to save time
    if process_this_frame:
        # Find all the faces and face encodings in the current frame of video
        face_locations = face_recognition.face_locations(rgb_small_frame)
        face_encodings = face_recognition.face_encodings(rgb_small_frame, face_locations)
​
        face_names = []
        for face_encoding in face_encodings:
            # See if the face is a match for the known face(s)
            matches = face_recognition.compare_faces(known_face_encodings, face_encoding, tolerance=0.5)
            # print("matchs:"+str(matches))
            name = "Un"
            # # If a match was found in known_face_encodings, just use the first one.
            # if True in matches:
            #     first_match_index = matches.index(True)
            #     name = known_face_names[first_match_index]
            # Or instead, use the known face with the smallest distance to the new face
            face_distances = face_recognition.face_distance(known_face_encodings, face_encoding)
            print("face_distances:" + str(face_distances))
            best_match_index = np.argmin(face_distances)
            if matches[best_match_index]:
                name = known_face_names[best_match_index]
​
            face_names.append(name)
​
    process_this_frame = not process_this_frame
    # Display the results
    for (top, right, bottom, left), name in zip(face_locations, face_names):
        # Scale back up face locations since the frame we detected in was scaled to 1/4 size
        # top *= 4
        # right *= 4
        # bottom *= 4
        # left *= 4
        # Draw a box around the face
        cv2.rectangle(frame, (left, top), (right, bottom), (0, 0, 255), 2)
        # Draw a label with a name below the face
        cv2.rectangle(frame, (left, bottom - 35), (right, bottom), (0, 0, 255), cv2.FILLED)
        font = cv2.FONT_HERSHEY_DUPLEX
        cv2.putText(frame, name, (left + 6, bottom - 6), font, 1.0, (255, 255, 255), 1)
    # Display the resulting image
    cv2.imshow('Video', frame)
    # Hit 'q' on the keyboard to quit!
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
# Release handle to the webcam
video_capture.release()
cv2.destroyAllWindows()

这里其实是dlib库的一个上层库，face_recognition,先pip install face_recognition,然后就能运行了。

至于原生的dlib写的人脸验证代码，我写成了一个服务，我们项目组也在一些地方使用了，对dlib的人脸识别进行了进一步的优化，达到了能工程落地，如果想要获得整个代码，请关注“暮秋梵星”，后台回复dlib，即可获得可以落地的人脸识别代码

今天的内容到此为止，后续有更加精彩的python小知识的文章更新，欢迎阅读。
对您能有哪怕一点帮助，就是我坚持写下去的动力。