基于阿里云人脸识别api实现人脸属性识别python源码+详细注释(佩戴口罩、帽子、年龄、表情等属性).zip

import tkinter as tk import cv2 import time from PIL import Image, ImageTk import json import oss2 from viapi.fileutils import FileUtils from aliyunsdkcore.client import AcsClient from aliyunsdkcore.acs_exception.exceptions import ClientException from aliyunsdkcore.acs_exception.exceptions import ServerException from aliyunsdkfacebody.request.v20191230.DetectMaskRequest import DetectMaskRequest from aliyunsdkfacebody.request.v20191230.RecognizeFaceRequest import RecognizeFaceRequest from aliyunsdkfacebody.request.v20191230.RecognizeExpressionRequest import RecognizeExpressionRequest # 阿里云的AccessKey和Secert(处于隐私和安全考虑，已省去) AccessKey = 'XXXXXXXXXXXXXXXXXXX' Secert = 'XXXXXXXXXXXXXXXXX' # 阿里云OSS对象存储的Bucket名字 BucketName = 'XXXXXXXX' # OpenCV分类器的完整路径 face_classfier_path = "haarcascade_frontalface_alt.xml" eye_classifer_path = "haarcascade_eye_tree_eyeglasses.xml" # 实例化阿里云的各种服务 auth = oss2.Auth(AccessKey, Secert) bucket = oss2.Bucket(auth, 'http://oss-cn-shanghai.aliyuncs.com', BucketName) file_utils = FileUtils(AccessKey, Secert) client = AcsClient(AccessKey, Secert, 'cn-shanghai') # 开启摄像头 camera = cv2.VideoCapture(0) # 定义openCV的分类器 fontface_classfier = cv2.CascadeClassifier(face_classfier_path) # 正脸识别 eye_classfier = cv2.CascadeClassifier(eye_classifer_path) # 眼睛识别 # left_eye_classfier = cv2.CascadeClassifier("haarcascade_lefteye_2splits.xml")#左眼识别 # 显示摄像头实时画面函数 ###################################################################################### def video_loop(): # 识别出人脸后要画的边框的颜色，RGB格式, color是一个不可增删的数组 color = (0, 255, 0) success, img = camera.read() # 从摄像头读取照片 if success: # 将当前桢图像转换成灰度图像 grey = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY) # 人脸检测，1.2和2分别为图片缩放比例和需要检测的有效点数 faceRects = fontface_classfier.detectMultiScale(grey, scaleFactor=1.2, minNeighbors=3, minSize=(32, 32)) eyesRects = eye_classfier.detectMultiScale(grey) if len(faceRects) > 0: # 大于0则检测到人脸 for faceRect in faceRects: # 单独框出每一张人脸 x, y, w, h = faceRect # 画出矩形框（绿色） cv2.rectangle(img, (x - 10, y - 10), (x + w + 10, y + h + 10), color, 2) if len(eyesRects) > 0: for (x, y, w, h) in eyesRects: # 画出眼睛的矩形方框（蓝色） frame = cv2.rectangle(img, (x, y), (x + w, y + h), (255, 0, 0), 2) cv2image = cv2.cvtColor(img, cv2.COLOR_BGR2RGBA) # 转换颜色从BGR到RGBA current_image = Image.fromarray(cv2image) # 将图像转换成Image对象 imgtk = ImageTk.PhotoImage(image=current_image) camera_panel.imgtk = imgtk camera_panel.config(image=imgtk) window.after(1, video_loop) # 更新右侧信息栏的函数，点击刷新按钮执行的函数 ################################################################################### def info_update(): # 用于存储图片的名字，图片对象，以及阿里云返回的url img_list = [] # 时间对象，用于给每张图片赋予不同的名字 time1 = time.localtime(time.time()) # 从摄像头读取照片 success, img = camera.read() if success: # 将当前桢图像转换成灰度图像 grey = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY) # 人脸检测，1.2和2分别为图片缩放比例和需要检测的有效点数 faceRects = fontface_classfier.detectMultiScale(grey, scaleFactor=1.2, minNeighbors=3, minSize=(32, 32)) # 眼睛检测 eyesRects = eye_classfier.detectMultiScale(grey) if len(faceRects) > 0: # 大于0则检测到人脸 for faceRect in faceRects: # 单独框出每一张人脸 x, y, w, h = faceRect # 通过人脸检测的矩形获取人脸图片 image_item = img[y - 50: y + h + 50, x - 50: x + w + 50] image_item = cv2.resize(image_item, (100, 100), interpolation=cv2.INTER_LINEAR) img_name = "images\\" + str(time1.tm_year) + str(time1.tm_mon) + str(time1.tm_mday) + str( time1.tm_hour) + str(time1.tm_min) + str(time1.tm_sec) + ".png" cv2.imwrite( "images\\" + str(time1.tm_year) + str(time1.tm_mon) + str(time1.tm_mday) + str(time1.tm_hour) + str( time1.tm_min) + str(time1.tm_sec) + ".png", image_item) # 上传到阿里云OSS服务的Bucket（该服务收费）同时获取图片的url bucket.put_object_from_file(img_name, img_name) oss_url = file_utils.get_oss_url(img_name, "png", True) # print(oss_url) img_info = [] img_info.append(img_name) img_info.append(image_item) img_info.append(oss_url) img_list.append(img_info) # 如果没获取到人脸但是获取到了眼睛，通过眼睛位置捕获人脸（可能重复捕获） elif len(eyesRects) > 0: for eyesRect in eyesRects: x, y, w, h = eyesRect image_item = img[y - 130: y + h + 130, x - 130: x + w + 130] image_item = cv2.resize(image_item, (100, 100), interpolation=cv2.INTER_AREA) img_name = "images\\" + str(time1.tm_year) + str(time1.tm_mon) + str(time1.tm_mday) + str( time1.tm_hour) + str(time1.tm_min) + str(time1.tm_sec) + ".png" cv2.imwrite( "images\\" + str(time1.tm_year) + str(time1.tm_mon) + str(time1.tm_mday) + str(time1.tm_hour) + str( time1.tm_min) + str(time1.tm_sec) + ".png", image_item) bucket.put_object_from_file(img_name, img_name) oss_url = file_utils.get_oss_url(img_name, "png", True) # print(oss_url) img_info = [] img_info.append(img_name) img_info.append(image_item) img_info.append(oss_url) img_list.append(img_info) # 销毁上一次的当前视频中的人脸信息 for item in frame_c_now.winfo_children(): item.destroy() frame_c_now.configure(height=0) # 更新右边的当前信息和历史记录 for item in img_list: # 0.是否佩戴口罩，1.颜色 Mask_info = DetectMask(item[2]) # 0.性别，1.年龄，2.颜值，3.眼镜，4.帽子 attributes = face_attributes(item[2]) # neutral（中性）、happiness（高兴）、surprise（惊讶）、sadness（伤心）、anger（生气）、disgust（厌恶）、fear（害怕）。 expression = detect_experssions(item[2]) # 更新右侧当前视频中人脸信息 item_Frame = tk.Frame(frame_c_now) item_Frame_right = tk.Frame(item_Frame) img_lable = tk.Label(item_Frame) mask_lable = tk.Label(item_Frame_right, text="是否佩戴口罩：" + Mask_info[0], fg=Mask_info[1]) sex_lable = tk.Label(item_Frame_right, text="性别：" + attributes[0] + " 年龄：" + str(attributes[1]) + " 颜值：" + str( attributes[2])) experssion_lable = tk.Label(item_Frame_right, text="表情：" + expression) glass_lable = tk.Label(item_Frame_right, text="是否佩戴眼镜：" + attributes[3]) hat_lable = tk.Label(item_Frame_right, text="是否佩戴帽子：" + attributes[4]) # 更新右侧历史记录人脸信息 item_Frame_history =