python 3利用Dlib 19.7实现摄像头人脸检测特征点标定

作者:TimeStamp 时间:2022-07-31 05:54:21 

Python 3 利用 Dlib 19.7 实现摄像头人脸检测特征点标定

0.引言

利用python开发,借助Dlib库捕获摄像头中的人脸,进行实时特征点标定;

python 3利用Dlib 19.7实现摄像头人脸检测特征点标定

图1 工程效果示例(gif)

python 3利用Dlib 19.7实现摄像头人脸检测特征点标定

图2 工程效果示例(静态图片)

(实现比较简单,代码量也比较少,适合入门或者兴趣学习。)

1.开发环境

python:3.6.3

dlib: 19.7

OpenCv, numpy


import dlib     # 人脸识别的库dlib
import numpy as np # 数据处理的库numpy
import cv2     # 图像处理的库OpenCv

2.源码介绍

其实实现很简单,主要分为两个部分:摄像头调用+人脸特征点标定

2.1 摄像头调用

介绍下opencv中摄像头的调用方法;

利用 cap = cv2.VideoCapture(0) 创建一个对象;

(具体可以参考官方文档)


# 2018-2-26
# By TimeStamp
# cnblogs: http://www.cnblogs.com/AdaminXie

"""
cv2.VideoCapture(), 创建cv2摄像头对象/ open the default camera

Python: cv2.VideoCapture() → <VideoCapture object>

Python: cv2.VideoCapture(filename) → <VideoCapture object>  
 filename – name of the opened video file (eg. video.avi) or image sequence (eg. img_%02d.jpg, which will read samples like img_00.jpg, img_01.jpg, img_02.jpg, ...)

Python: cv2.VideoCapture(device) → <VideoCapture object>
 device – id of the opened video capturing device (i.e. a camera index). If there is a single camera connected, just pass 0.

"""
cap = cv2.VideoCapture(0)

"""
cv2.VideoCapture.set(propId, value),设置视频参数;

propId:
 CV_CAP_PROP_POS_MSEC Current position of the video file in milliseconds.
 CV_CAP_PROP_POS_FRAMES 0-based index of the frame to be decoded/captured next.
 CV_CAP_PROP_POS_AVI_RATIO Relative position of the video file: 0 - start of the film, 1 - end of the film.
 CV_CAP_PROP_FRAME_WIDTH Width of the frames in the video stream.
 CV_CAP_PROP_FRAME_HEIGHT Height of the frames in the video stream.
 CV_CAP_PROP_FPS Frame rate.
 CV_CAP_PROP_FOURCC 4-character code of codec.
 CV_CAP_PROP_FRAME_COUNT Number of frames in the video file.
 CV_CAP_PROP_FORMAT Format of the Mat objects returned by retrieve() .
 CV_CAP_PROP_MODE Backend-specific value indicating the current capture mode.
 CV_CAP_PROP_BRIGHTNESS Brightness of the image (only for cameras).
 CV_CAP_PROP_CONTRAST Contrast of the image (only for cameras).
 CV_CAP_PROP_SATURATION Saturation of the image (only for cameras).
 CV_CAP_PROP_HUE Hue of the image (only for cameras).
 CV_CAP_PROP_GAIN Gain of the image (only for cameras).
 CV_CAP_PROP_EXPOSURE Exposure (only for cameras).
 CV_CAP_PROP_CONVERT_RGB Boolean flags indicating whether images should be converted to RGB.
 CV_CAP_PROP_WHITE_BALANCE_U The U value of the whitebalance setting (note: only supported by DC1394 v 2.x backend currently)
 CV_CAP_PROP_WHITE_BALANCE_V The V value of the whitebalance setting (note: only supported by DC1394 v 2.x backend currently)
 CV_CAP_PROP_RECTIFICATION Rectification flag for stereo cameras (note: only supported by DC1394 v 2.x backend currently)
 CV_CAP_PROP_ISO_SPEED The ISO speed of the camera (note: only supported by DC1394 v 2.x backend currently)
 CV_CAP_PROP_BUFFERSIZE Amount of frames stored in internal buffer memory (note: only supported by DC1394 v 2.x backend currently)

value: 设置的参数值/ Value of the property
"""
cap.set(3, 480)

"""
cv2.VideoCapture.isOpened(), 检查摄像头初始化是否成功 / check if we succeeded
返回true或false
"""
cap.isOpened()

"""
cv2.VideoCapture.read([imgage]) -> retval,image, 读取视频 / Grabs, decodes and returns the next video frame
返回两个值:
 一个是布尔值true/false,用来判断读取视频是否成功/是否到视频末尾
 图像对象,图像的三维矩阵
"""
flag, im_rd = cap.read()

2.2 人脸特征点标定

调用预测器“shape_predictor_68_face_landmarks.dat”进行68点标定,这是dlib训练好的模型,可以直接调用进行人脸68个人脸特征点的标定;

具体可以参考我的另一篇博客(python3利用Dlib19.7实现人脸68个特征点标定); 

2.3 源码

实现的方法比较简单:

利用 cv2.VideoCapture() 创建摄像头对象,然后利用 flag, im_rd = cv2.VideoCapture.read() 读取摄像头视频,im_rd就是视频中的一帧帧图像;

然后就类似于单张图像进行人脸检测,对这一帧帧的图像im_rd利用dlib进行特征点标定,然后绘制特征点;

你可以按下s键来获取当前截图,或者按下q键来退出摄像头;

# 2018-2-26


# By TimeStamp
# cnblogs: http://www.cnblogs.com/AdaminXie
# github: https://github.com/coneypo/Dlib_face_detection_from_camera

import dlib           #人脸识别的库dlib
import numpy as np       #数据处理的库numpy
import cv2           #图像处理的库OpenCv

# dlib预测器
detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor('shape_predictor_68_face_landmarks.dat')

# 创建cv2摄像头对象
cap = cv2.VideoCapture(0)

# cap.set(propId, value)
# 设置视频参数,propId设置的视频参数,value设置的参数值
cap.set(3, 480)

# 截图screenshoot的计数器
cnt = 0

# cap.isOpened() 返回true/false 检查初始化是否成功
while(cap.isOpened()):

# cap.read()
 # 返回两个值:
 #  一个布尔值true/false,用来判断读取视频是否成功/是否到视频末尾
 #  图像对象,图像的三维矩阵
 flag, im_rd = cap.read()

# 每帧数据延时1ms,延时为0读取的是静态帧
 k = cv2.waitKey(1)

# 取灰度
 img_gray = cv2.cvtColor(im_rd, cv2.COLOR_RGB2GRAY)

# 人脸数rects
 rects = detector(img_gray, 0)

#print(len(rects))

# 待会要写的字体
 font = cv2.FONT_HERSHEY_SIMPLEX

# 标68个点
 if(len(rects)!=0):
   # 检测到人脸
   for i in range(len(rects)):
     landmarks = np.matrix([[p.x, p.y] for p in predictor(im_rd, rects[i]).parts()])

for idx, point in enumerate(landmarks):
       # 68点的坐标
       pos = (point[0, 0], point[0, 1])

# 利用cv2.circle给每个特征点画一个圈,共68个
       cv2.circle(im_rd, pos, 2, color=(0, 255, 0))

# 利用cv2.putText输出1-68
       cv2.putText(im_rd, str(idx + 1), pos, font, 0.2, (0, 0, 255), 1, cv2.LINE_AA)
   cv2.putText(im_rd, "faces: "+str(len(rects)), (20,50), font, 1, (0, 0, 255), 1, cv2.LINE_AA)
 else:
   # 没有检测到人脸
   cv2.putText(im_rd, "no face", (20, 50), font, 1, (0, 0, 255), 1, cv2.LINE_AA)

# 添加说明
 im_rd = cv2.putText(im_rd, "s: screenshot", (20, 400), font, 0.8, (255, 255, 255), 1, cv2.LINE_AA)
 im_rd = cv2.putText(im_rd, "q: quit", (20, 450), font, 0.8, (255, 255, 255), 1, cv2.LINE_AA)

# 按下s键保存
 if (k == ord('s')):
   cnt+=1
   cv2.imwrite("screenshoot"+str(cnt)+".jpg", im_rd)

# 按下q键退出
 if(k==ord('q')):
   break

# 窗口显示
 cv2.imshow("camera", im_rd)

# 释放摄像头
cap.release()

# 删除建立的窗口
cv2.destroyAllWindows()

如果对您有帮助,欢迎在GitHub上star本项目。

来源:https://www.cnblogs.com/AdaminXie/p/8472743.html

标签:python3,Dlib19.7,摄像头,人脸检测
0
投稿

猜你喜欢

  • 用javascript连接access数据库的方法

    2024-01-15 11:11:51
  • python进度条库tqdm的基本操作方法

    2021-03-02 08:31:22
  • javascript 操作文件 实现方法小结

    2024-04-22 12:48:27
  • pytorch 彩色图像转灰度图像实例

    2023-08-02 17:28:37
  • Django 添加静态文件的两种实现方法(必看篇)

    2021-09-03 23:53:58
  • SQL Server 常用函数使用方法小结

    2024-01-13 11:39:32
  • Google Chrome CSS选择器速度测试比较

    2008-10-06 13:24:00
  • python文件和目录操作方法大全(含实例)

    2021-11-11 14:10:29
  • python模块和函数帮助文档快速查看方法示例

    2023-03-31 00:24:04
  • 精简版的MySQL制作步骤

    2011-03-08 09:52:00
  • Python参数传递对象的引用原理解析

    2023-04-23 18:33:57
  • 详解Python如何实现批量为PDF添加水印

    2022-06-20 23:33:58
  • 搭建websocket消息推送服务,必须要考虑的几个问题

    2023-11-28 12:23:45
  • 一个可应用在ASP 标记加密文件的MD5的DLL组件

    2008-04-12 07:21:00
  • js DNA动态序列比对代码

    2024-04-16 10:41:26
  • css设计小巧三条

    2008-01-21 13:04:00
  • Python‘==‘ 及 ‘is‘相关原理解析

    2021-10-11 15:16:59
  • Wordpress 相册插件 NextGEN-Gallery 添加目录将中文转为拼音的解决办法

    2023-09-04 23:12:41
  • python多进程下的生产者和消费者模型

    2022-05-30 02:37:07
  • python ipset管理 增删白名单的方法

    2021-02-10 17:38:19
  • asp之家 网络编程 m.aspxhome.com