准备阶段

  1. 准备三个目录 一个放 识别好的人脸标本，一个放需要识别的照片，一个放识别后的结果

比如：

python利用opencv实现本地图片的识别与结果存储_opencv

  1. 需要安装opencv包、dlib包，以及下载人脸数据，参照：​​python利用opencv读取摄像头数据​​
  2. 另外还需要安装人脸比对的包 
实现步骤
  1. 读取识别好的人脸标本，并解析
  2. 读取需要检测的人脸图片，解析并与标本比对
  3. 比对成功则绘制绿框，标出姓名
代码实现
# This is a sample Python script.
import os.path

# Press Shift+F10 to execute it or replace it with your code.
# Press Double Shift to search everywhere for classes, files, tool windows, actions, and settings.
import cv2 as cv
import dlib
import numpy as np
import face_recognition

# 加载人脸检测器和关键点检测器
detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor("shape_predictor_68_face_landmarks.dat")

# 已知人脸存储路径
known_face_path = 'faces/gugu/'
# 需要检测的人脸存储路径
unknown_face_path = 'faces/gys/'
# 记录结果
record_face_path = 'faces/record/'


def recognize_image():
    # 读取已知人脸，并解析
    known_face_encodings = []
    known_face_names = []
    print('开始加载已知人脸数据...')
    for root, dirs, files in os.walk(known_face_path):
        total_process = len(files)
        process = 0
        for f, file in enumerate(files):
            img = face_recognition.load_image_file(known_face_path + file)
            face_encodings = face_recognition.face_encodings(img, model='cnn')
            if len(face_encodings) > 0:
                known_face_encodings.append(face_encodings[0])
                known_face_names.append(file[0:file.index('.')])
            process += 1
            print('----> 已加载' + str(process / total_process * 100 // 1) + '%')
    print('数据加载完成！')

    # 识别人脸数据
    if not os.path.exists(record_face_path):
        os.makedirs(record_face_path)

    for root, dirs, files in os.walk(unknown_face_path):
        total_process = len(files)
        process = 0
        for f, file in enumerate(files):
            cv_img = cv.imdecode(np.fromfile(unknown_face_path + file, dtype=np.uint8), cv.IMREAD_COLOR)

            # img = face_recognition.load_image_file(unknown_face_path + file)
            img = cv.cvtColor(cv_img, cv.COLOR_BGR2RGB)
            locations = face_recognition.face_locations(img)
            face_encodings = face_recognition.face_encodings(img, locations, model='cnn')
            for (top, right, bottom, left), face_encoding in zip(locations, face_encodings):
                matches = face_recognition.compare_faces(known_face_encodings, face_encoding, .5)
                if True in matches:
                    index = matches.index(True)
                    name = known_face_names[index]
                    # 绘制矩形框
                    cv.rectangle(cv_img, (left, top, right - left, bottom - top), (0, 255, 0), 2)
                    cv.putText(cv_img, name, (left + 10, bottom - 10), cv.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255),  2)
            cv.imencode('.jpg', cv_img, [cv.IMWRITE_JPEG_QUALITY, 100])[1].tofile(record_face_path + '副本_' + file)

            process += 1
            print('----> 已识别' + str(process / total_process * 100 // 1) + '%')


# Press the green button in the gutter to run the script.
if __name__ == '__main__':
    print_hi('PyCharm')
    recognize_image()
已知问题：如果姓名是中文，绘制到图片上，会以问号的形式打印

注意：matches = face_recognition.compare_faces(known_face_encodings, face_encoding, .5) 中需要调整比较距离tolerance的值，以达到需要的效果

输出结果
python利用opencv实现本地图片的识别与结果存储_opencv_02