人脸检索 M:N(视频，摄像头)，调用百度API-编程知识

创建百度智能云账号

代码部分：

创建百度智能云账号

网址：

百度智能云-云智一体深入产业

点击导航栏中的产品，再选择人脸与人体

再选择人脸搜索

进入后，可以先去领取免费资源，如果不领取，后面是无法调用api的，建议是把所有的免费资源都领取，我是全部都领取了。

然后就是创建应用，按照要求填写就好。

创建成功后，可以在应用列表看到自己创建好的应用。下图用红色方框圈起来的在后续代码中需要填写。

接着我们可以创建我们的可视化人脸库。

按照要求创建即可，下图是我已经创建成功的。

再点进自己创建好的人脸库，创建分组，这里我已经创建好了四个分组，接着就是进入分组，创建用户，添加图片

这一切完成后，就有了自己的人像数据库。

代码部分：

完整代码：

import cv2
from aip import AipFace
from io import BytesIO
import base64
from PIL import Image,ImageDraw, ImageFont
import threading
import numpy as np
import imutils
import time# 在图像上绘制文本
def cv2ImgAddText(img, text, left, top, textColor=(0, 255, 0), textSize=20):if (isinstance(img, np.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("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(np.asarray(img), cv2.COLOR_RGB2BGR)# 将图像格式转换为base64格式
def frame2base64(frame):img = Image.fromarray(frame)  # 将每一帧转为Imageoutput_buffer = BytesIO()  # 创建一个BytesIOimg.save(output_buffer, format='JPEG')  # 写入output_bufferbyte_data = output_buffer.getvalue()  # 在内存中读取base64_data = base64.b64encode(byte_data)  # 转为BASE64return base64_data  # 转码成功 返回base64编码def process(image,ls):""" 调用人脸检测 """""" 如果有可选参数 """# 换成自己的百度appidAPP_ID = ''API_KEY = ''SECRET_KEY = ''groupIdList = ''imageType = 'BASE64'base64 = frame2base64(image)base64 = str(base64,'UTF-8')options = {}options["max_face_num"] = 10options["face_type"] = "LIVE"options["match_threshold"] = 0#options["liveness_control"] = "NORMAL"client = AipFace(APP_ID, API_KEY, SECRET_KEY)""" 如果有可选参数 """options = {}options["max_face_num"] = 10options["match_threshold"] = 0options["quality_control"] = "LOW"options["liveness_control"] = "LOW"options["max_user_num"] = 7start3 = time.time()json1 = client.multiSearch(base64, imageType, groupIdList, options)print("*******:", time.time() - start3)print("json1:", json1)if json1['error_msg'] == 'SUCCESS':face_num = json1['result']['face_num']for i in range(face_num):x = max(int(json1['result']['face_list'][i]['location']['left']), 0)y = max(int(json1['result']['face_list'][i]['location']['top']), 0)width = int(json1['result']['face_list'][i]['location']['width'])height = int(json1['result']['face_list'][i]['location']['height'])cv2.rectangle(image, (x, y), (x + width, y + height), (0, 0, 255), 2)print("$$$$$$$$$$$$$$$$$$$$$$$$$")print("json1:", json1)if json1['result']['face_list'][i]['user_list'][0]['score']>70:print(json1['result']['face_list'][i]['user_list'][0]['user_id'])image = cv2ImgAddText(image, json1['result']['face_list'][i]['user_list'][0]['user_id'],max(x - 20, 0), max(y - 20, 0), (255, 0, 0), 60)else:image = cv2ImgAddText(image, "访客",max(x - 20, 0), max(y - 20, 0), (255, 0, 0), 60)ls.append(image)else:print("************************")print("没有检测到人脸：", json1)def main():# 使用摄像头检测# video_capture = cv2.VideoCapture(0)# 使用本地视频文件检测video_capture = cv2.VideoCapture('test6.mp4')fps_time =0f = 0start = time.time()while True:f += 1ls = []if f == 30:print("时间到：", time.time()-start)breakret, frame = video_capture.read()# t = threading.Thread(target=process, args=(frame, ls))  # 使用线程# t.start()# t.join()process(frame, ls)frame = ls[0] if ls else frameframe = imutils.resize(frame, width=400)frame = cv2.putText(frame, '%d, FPS: %f' % (f, 1.0 / (time.time() - fps_time)),(10, 20), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 1)fps_time = time.time()cv2.imshow('wx', frame)if cv2.waitKey(1) & 0xFF == ord('Q'):cv2.destroyAllWindows()video_capture.release()break
if __name__ == "__main__":main()

后续写了一点注册，更新人脸数据库的代码：

from aip import AipFace
from PIL import Image
from io import BytesIO
import base64def frame2base64(image_path):# img = Image.fromarray(frame)  # 将每一帧转为Imageimg = Image.open(image_path)output_buffer = BytesIO()  # 创建一个BytesIOimg.save(output_buffer, format='JPEG')  # 写入output_bufferbyte_data = output_buffer.getvalue()  # 在内存中读取base64_data = base64.b64encode(byte_data)  # 转为BASE64return base64_data  # 转码成功 返回base64编码def user_register(image,groupId, userId):""" 你的 APPID AK SK """APP_ID = ''API_KEY = ''SECRET_KEY = ''client = AipFace(APP_ID, API_KEY, SECRET_KEY)# 将图片转换为 base64 类型image = frame2base64(image)image = str(image, 'UTF-8')# 图片类型imageType = "BASE64"# 分组groupId = groupId# 用户iduserId = userId""" 调用人脸注册 """json1 = client.addUser(image, imageType, groupId, userId)while json1['error_msg'] == 'Open api qps request limit reached':json1 = client.addUser(image, imageType, groupId, userId)print("jason1:", json1)# """ 如果有可选参数 """# options = {}# options["user_info"] = "user's info"# options["quality_control"] = "NORMAL"# options["liveness_control"] = "LOW"# options["action_type"] = "REPLACE"## """ 带参数调用人脸注册 """# client.addUser(image, imageType, groupId, userId, options)def user_update(image, groupId,userId):""" 你的 APPID AK SK """APP_ID = ''API_KEY = ''SECRET_KEY = ''client = AipFace(APP_ID, API_KEY, SECRET_KEY)# 将图片转换为 base64 类型image = frame2base64(image)image = str(image, 'UTF-8')imageType = "BASE64"# 分组IdgroupId = groupId# 用户IduserId = userId""" 调用人脸更新 """jason2 = client.updateUser(image, imageType, groupId, userId)print("jason2:", jason2)# """ 如果有可选参数 """# options = {}# options["user_info"] = "user's info"# options["quality_control"] = "NORMAL"# options["liveness_control"] = "LOW"# options["action_type"] = "REPLACE"## """ 带参数调用人脸更新 """# client.updateUser(image, imageType, groupId, userId, options)if __name__ == "__main__":image = "L1.JPG"groupId = "group"userId = "2020"user_register(image, groupId, userId)

百度也提供了详细的 Python-SDK 供大家学习使用。