Python笔记-OpenCV图像处理和人脸识别

一.简介

1.OpenCV：OpenCV的全称是Open Source Computer Vision Library，是一个开源的跨平台的计算机视觉库。可以运行在Linux、Windows、Android和macOS操作系统上，帮助人们快速构建复杂的视觉应用程序。

2.计算机视觉：计算机视觉（Computer Vision）就是利用计算机来处理图像，将来自静止或摄像机的数据转换成新的表示方式，获得我们想要的信息。

二.编程步骤

1.环境配置

2.读取图片

3.灰度转换

4.修改尺寸

5.绘制矩形

6.人脸检测

7.检测多个

8.视频检测

9.拍照保存

10.训练数据

11.人脸识别

三.简单实例

1.环境配置

python下编程，先安装库

pip install opencv-python
pip install opencv-contrib-python

通过pip下载的只是阉割版的，部分功能不全，要实现人脸识别还需在官网Home - OpenCV下载完整的库(官网有点慢，可以多等等)（在人脸检测的部分需要手动添加库的路径，我的是Mac所以路径是/Users/....）（主要用到haarcascade_frontalface_default.xml、haarcascade_frontalface_alt2.xml

）

2.读取图片

#导入cv模块
import cv2 as cv
#读取图片
img = cv.imread('face1.jpg')#图片路径
#显示图片
cv.imshow('read_img',img)
#等待
cv.waitKey(0)
#释放内存
cv.destroyAllWindows()

3.灰度转换

#导入cv模块
import cv2 as cv
#读取图片
img = cv.imread('face1.jpg')
#灰度转换
gray_img = cv.cvtColor(img,cv.COLOR_BGR2GRAY)
#显示灰度图片
cv.imshow('gray',gray_img)
#保存灰度图片
cv.imwrite('gray_face1.jpg',gray_img)
#显示图片
cv.imshow('read_img',img)
#等待
cv.waitKey(0)
#释放内存
cv.destroyAllWindows()

4.修改尺寸

#导入cv模块
import cv2 as cv
#读取图片
img = cv.imread('face1.jpg')
#修改尺寸
resize_img = cv.resize(img,dsize=(200,200))
#显示原图
cv.imshow('img',img)
#显示修改后的
cv.imshow('resize_img',resize_img)
#打印原图尺寸大小
print('未修改：',img.shape)
#打印修改后的大小
print('修改后：',resize_img.shape)
#等待
while True:if ord('q') == cv.waitKey(0):break
#释放内存
cv.destroyAllWindows()

5.绘制矩形

#导入cv模块
import cv2 as cv
#读取图片
img = cv.imread('face1.jpg')
#坐标
x,y,w,h = 100,100,100,100
#绘制矩形
cv.rectangle(img,(x,y,x+w,y+h),color=(0,0,255),thickness=1)
#绘制圆形
cv.circle(img,center=(x+w,y+h),radius=100,color=(255,0,0),thickness=5)
#显示
cv.imshow('re_img',img)
while True:if ord('q') == cv.waitKey(0):break
#释放内存
cv.destroyAllWindows()

6.人脸检测

# 导入cv模块
import cv2 as cv
# 检测函数
def face_detect_demo():gary = cv.cvtColor(img, cv.COLOR_BGR2GRAY)face_detect = cv.CascadeClassifier('/Users/huangjw/Downloads/opencv-4.x/data/haarcascades/haarcascade_frontalface_alt2.xml')face = face_detect.detectMultiScale(gary, 1.01, 5, 0, (100, 100), (300, 300))for x, y, w, h in face:cv.rectangle(img, (x, y), (x + w, y + h), color=(0, 0, 255), thickness=2)cv.imshow('result', img)# 读取图像
img = cv.imread('face1.jpg')
#img = cv.imread('/Users/huangjw/Downloads/mycodetest/opencv/data/hjw/2.hjw.jpg')
# 检测函数
face_detect_demo()
# 等待
while True:if ord('q') == cv.waitKey(0):break
# 释放内存
cv.destroyAllWindows()

7.检测多个(同时检测多个人脸)

# 导入cv模块
import cv2 as cv# 检测函数
def face_detect_demo():gary = cv.cvtColor(img, cv.COLOR_BGR2GRAY)face_detect = cv.CascadeClassifier( '/Users/huangjw/Downloads/opencv-4.x/data/haarcascades/haarcascade_frontalface_default.xml')face = face_detect.detectMultiScale(gary)for x, y, w, h in face:cv.rectangle(img, (x, y), (x + w, y + h), color=(0, 0, 255), thickness=2)cv.imshow('result', img)# 读取图像
img = cv.imread('face2.jpg')
# 检测函数
face_detect_demo()
# 等待
while True:if ord('q') == cv.waitKey(0):break
# 释放内存
cv.destroyAllWindows()

8.视频检测

# 导入cv模块
import cv2 as cv# 检测函数
def face_detect_demo(img):gary = cv.cvtColor(img, cv.COLOR_BGR2GRAY)face_detect = cv.CascadeClassifier('/Users/huangjw/Downloads/opencv-4.x/data/haarcascades/haarcascade_frontalface_default.xml')face = face_detect.detectMultiScale(gary)for x, y, w, h in face:cv.rectangle(img, (x, y), (x + w, y + h), color=(0, 0, 255), thickness=2)cv.imshow('result', img)# 读取摄像头
cap = cv.VideoCapture(0)
# 循环
while True:flag, frame = cap.read()if not flag:breakface_detect_demo(frame)if ord('q') == cv.waitKey(1):break
# 释放内存
cv.destroyAllWindows()
# 释放摄像头
cap.release()

9.拍照保存

调用摄像头拍照，按下‘s’键拍照,将照片保存在指定路径，空格键退出程序

# 导入模块
import cv2# 摄像头
cap = cv2.VideoCapture(0)#falg = 1
num = 1while (cap.isOpened()):  # 检测是否在开启状态ret_flag, Vshow = cap.read()  # 得到每帧图像cv2.imshow("Capture_Test", Vshow)  # 显示图像k = cv2.waitKey(1) & 0xFF  # 按键判断if k == ord('s'):  # 保存cv2.imwrite("/Users/huangjw/Downloads/mycodetest/opencv/data/four/" + str(num) + ".hjc" + ".jpg", Vshow)print("success to save" + str(num) + ".jpg")num += 1elif k == ord(' '):  # 退出break
# 释放摄像头
cap.release()
# 释放内
cv2.destroyAllWindows()

10.训练数据

将所需人脸识别的图片训练，生成yml文件，以供人脸识别

import os
import cv2
import sys
from PIL import Image
import numpy as npdef getImageAndLabels(path):facesSamples = []ids = []imagePaths = [os.path.join(path, f) for f in os.listdir(path)]# 检测人脸face_detector = cv2.CascadeClassifier('/Users/huangjw/Downloads/opencv-4.x/data/haarcascades/haarcascade_frontalface_alt2.xml')# 打印数组imagePathsprint('数据排列：', imagePaths)# 遍历列表中的图片for imagePath in imagePaths:# 打开图片,黑白化PIL_img = Image.open(imagePath).convert('L')# 将图像转换为数组，以黑白深浅# PIL_img = cv2.resize(PIL_img, dsize=(400, 400))img_numpy = np.array(PIL_img, 'uint8')# 获取图片人脸特征faces = face_detector.detectMultiScale(img_numpy)# 获取每张图片的id和姓名id = int(os.path.split(imagePath)[1].split('.')[0])# 预防无面容照片for x, y, w, h in faces:ids.append(id)facesSamples.append(img_numpy[y:y + h, x:x + w])# 打印脸部特征和id# print('fs:', facesSamples)print('id:', id)# print('fs:', facesSamples[id])print('fs:', facesSamples)# print('脸部例子：',facesSamples[0])# print('身份信息：',ids[0])return facesSamples, idsif __name__ == '__main__':# 图片路径path = './data/four/'# 获取图像数组和id标签数组和姓名faces, ids = getImageAndLabels(path)# 获取训练对象recognizer = cv2.face.LBPHFaceRecognizer_create()# recognizer.train(faces,names)#np.array(ids)recognizer.train(faces, np.array(ids))# 保存文件recognizer.write('trainer/four.yml')# save_to_file('names.txt',names)

11.人脸识别

通过之前训练好的数据，对所需识别的图像、视频等加以分类（一下程序需要修改训练数据路径、训练图片路径(获取图片名字)、识别图像的方式）

import cv2
import numpy as np
import os
# coding=utf-8
import urllib
import urllib.request
import hashlib# 加载训练数据集文件
recogizer = cv2.face.LBPHFaceRecognizer_create()
recogizer.read('/Users/huangjw/Downloads/mycodetest/opencv/trainer/trainer.yml')
# recogizer.read('trainer/trainer.yml')
names = []
warningtime = 0# 准备识别的图片
def face_detect_demo(img):gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)  # 转换为灰度# face_detector = cv2.CascadeClassifier(#   '/Users/huangjw/Downloads/opencv-4.x/data/haarcascades/haarcascade_frontalface_alt2.xml')# 加载分类器（opencv已经训练好了）face_detector = cv2.CascadeClassifier('/Users/huangjw/Downloads/opencv-4.x/data/haarcascades/haarcascade_frontalface_default.xml')face = face_detector.detectMultiScale(gray, 1.1, 5, cv2.CASCADE_SCALE_IMAGE, (100, 100), (500, 500))# face=face_detector.detectMultiScale(gray)for x, y, w, h in face:cv2.rectangle(img, (x, y), (x + w, y + h), color=(0, 0, 255), thickness=2)cv2.circle(img, center=(x + w // 2, y + h // 2), radius=w // 2, color=(0, 255, 0), thickness=1)# 人脸识别ids, confidence = recogizer.predict(gray[y:y + h, x:x + w])# print('标签id:',ids,'置信评分：', confidence)if confidence > 70:global warningtimewarningtime += 1if warningtime > 100:# 发送警报# warning()print("陌生人")warningtime = 0cv2.putText(img, 'unkonw', (x + 10, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.75, (0, 255, 0), 1)else:#print(ids - 1)cv2.putText(img, str(names[ids - 1]), (x + 10, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.75, (0, 255, 0), 1)cv2.imshow('result', img)# print('bug:',ids)def name():path = '/Users/huangjw/Downloads/mycodetest/opencv/data/jm/'# names = []imagePaths = [os.path.join(path, f) for f in os.listdir(path)]print('数据排列：', imagePaths)for imagePath in imagePaths:name = str(os.path.split(imagePath)[1].split('.', 2)[1])names.append(name)#print(names)# , cv.cvtColor(img, cv.COLOR_BGR2GRAY)
# 读取视频
#cap = cv2.VideoCapture('1.mp4')# 读取摄像头
#cap = cv2.VideoCapture(0)name()
print(names)
# names.reverse()  # 队列反转
# print(names)
while True:flag, frame = cap.read()if not flag:breakface_detect_demo(frame)if ord(' ') == cv2.waitKey(10):break
cv2.destroyAllWindows()
cap.release()
# print(names)