python 人脸识别demo

使用python第三方模块face_recognition实现人脸识别，并根据已命名的图片把名字显示在屏幕上。

1. 安装模块
   需要安装opencv，face_recognition,face_recognition模块需要先安装dlib，而dlib需要先安装cmake和boost
   所以按顺序安装

pip install cmake
pip install boost
pip install dlib
pip install face_recognition
pip install opencv-python

如果未安装pip等需要工具，请自行百度。
如果速度慢，可以在命令后加 -i [国内源]
如

pip install opencv-python -i https://mirrors.aliyun.com/pypi/simple

我用的三个国内源：
阿里云 https://mirrors.aliyun.com/pypi/simple
清华大学 https://pypi.tuna.tsinghua.edu.cn/simple
中国科技大学 https://pypi.mirrors.ustc.edu.cn/simple

2. 代码实现

import face_recognition
import cv2
import os

import numpy
from PIL import Image, ImageDraw, ImageFont


def cv2ImgAddText(img, text, left, top, textColor=(0, 255, 0), textSize=20):
    if (isinstance(img, numpy.ndarray)):  # 判断是否OpenCV图片类型
        img = Image.fromarray(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
    # 创建一个可以在给定图像上绘图的对象
    draw = ImageDraw.Draw(img)
    # 字体的格式
    fontStyle = ImageFont.truetype(
        "font/simsun.ttc", textSize, encoding="utf-8")
    # 绘制文本
    draw.text((left, top), text, textColor, font=fontStyle)
    # 转换回OpenCV格式
    return cv2.cvtColor(numpy.asarray(img), cv2.COLOR_RGB2BGR)


def face(path):
    # 存储知道人名列表
    global face_locations, face_names
    known_names = []
    # 存储知道的特征值
    known_encodings = []
    for image_name in os.listdir(path):
        print(path + image_name)
        load_image = face_recognition.load_image_file(path + image_name)  # 加载图片
        image_face_encoding = face_recognition.face_encodings(load_image)[0]  # 获得128维特征值
        known_names.append(image_name.split(".")[0])
        known_encodings.append(image_face_encoding)
    print(known_encodings)

    # 打开摄像头，0表示内置摄像头
    video_capture = cv2.VideoCapture(1)
    process_this_frame = True
    while True:
        ret, frame = video_capture.read()
        # opencv的图像是BGR格式的，而我们需要是的RGB格式的，因此需要进行一个转换。
        rgb_frame = frame[:, :, ::-1]
        if process_this_frame:
            face_locations = face_recognition.face_locations(rgb_frame)  # 获得所有人脸位置
            face_encodings = face_recognition.face_encodings(rgb_frame, face_locations)  # 获得人脸特征值
            face_names = []  # 存储出现在画面中人脸的名字
            for face_encoding in face_encodings:
                matches = face_recognition.compare_faces(known_encodings, face_encoding, tolerance=0.3)
                if True in matches:
                    first_match_index = matches.index(True)
                    name = known_names[first_match_index]
                else:
                    name = "unknown"
                face_names.append(name)

        process_this_frame = not process_this_frame

        # 将捕捉到的人脸显示出来
        for (top, right, bottom, left), name in zip(face_locations, face_names):
            cv2.rectangle(frame, (left, top), (right, bottom), (0, 0, 255), 2)  # 画人脸矩形框
            # 加上人名标签
            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)
            frame=cv2ImgAddText(frame, name, left + 32, bottom - 32, textColor=(255, 255, 255), textSize=32)

        cv2.imshow('frame', frame)
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break

    video_capture.release()
    cv2.destroyAllWindows()


if __name__ == '__main__':
    face("./images/")  # 存放已知图像路径

其中，face（）是识别的方法，cv2ImgAddText（）是解决图片添加文字时的中文乱码问题。
matches = face_recognition.compare_faces(known_encodings, face_encoding, tolerance=0.3）是对数据进行比对，tolerance越小，进度越高，一般0.6是性能最好的。

3. 效果