opencv3+python实现人脸追踪

转载于 2021-10-02 11:35:58 发布 · 257 阅读

0 ·

CC 4.0 BY-SA版权

原文链接：https://www.jianshu.com/p/b50dc83503d3

文章标签：

#opencv #python

本文档详细介绍了如何在Ubuntu上安装Python 2.7和OpenCV 3.3.0，包括依赖安装、CMake配置、编译和安装步骤。在编译OpenCV时遇到CMAKE错误，通过查找FFmpeg头文件解决问题。最后，安装了numpy库，并展示了一个人脸识别的Python示例代码。

转自：https://www.jianshu.com/p/b50dc83503d3

Install Python

python 2.7

Install OpenCV 3.3.0

Script

$ vim opencv.sh3.3.0-20180429-yuan

#!/bin/bash

echo "========================================"
echo "Install OpenCV 3.3.0"
echo "========================================"

echo "-> install dependency"
echo "========================================"
#sudo apt-get install build-essential
#sudo apt-get install cmake git libgtk2.0-dev pkg-config libavcodec-dev libavformat-dev libswscale-dev
#sudo apt-get install python-dev python-numpy libtbb2 libtbb-dev libjpeg-dev libpng-dev libtiff-dev libjasper-dev libdc1394-22-dev

echo "-> cmake"
echo "========================================"
#cmake ..

echo "-> make"
echo "========================================"
#make

echo "-> install"
echo "========================================"
#sudo make install

echo "-> check"
echo "========================================"
pkg-config --modversion opencv

echo "========================================"
echo "Installation finished"

Update CMake

可行：
sudo apt-get install software-properties-mmon 
sudo add-apt-repository ppa:george-edison55/cmake-3.x 
sudo apt-get update  
sudo apt-get upgrade
//sudo apt-get install cmake
cmake --version

CMake Error:

[ 24%] Building CXX object /home/xy2/tools/opencv-3.3.0/modules/videoio/src/cap_ffmpeg_impl.hpp:1557:21: 
error: ‘CODEC_FLAG_GLOBAL_HEADER’ was not declared in this scope
         c->flags |= CODEC_FLAG_GLOBAL_HEADER;
                     ^

Solution:

https://stackoverflow.com/questions/46884682/error-in-building-opencv-with-ffmpeg
My solution is to grep the missing defines (2 in total) from FFmpeg by using grep -r which leads to the following code found in libavcodec/avcodec.h:

#define AV_CODEC_FLAG_GLOBAL_HEADER (1 << 22)
#define CODEC_FLAG_GLOBAL_HEADER AV_CODEC_FLAG_GLOBAL_HEADER
#define AVFMT_RAWPICTURE 0x0020
Copy and paste it to the top of:

opencv-3.3.0/modules/videoio/src/cap_ffmpeg_impl.hpp
Compile and everything works even with the latest sources

$ ag "CODEC_FLAG_GLOBAL_HEADER" /usr/include/
/usr/include/x86_64-linux-gnu/libavcodec/avcodec.h
784:#define AV_CODEC_FLAG_GLOBAL_HEADER   (1 << 22)
999:#define CODEC_FLAG_GLOBAL_HEADER  AV_CODEC_FLAG_GLOBAL_HEADER

$ vim /home/xy2/tools/opencv-3.3.0/modules/videoio/src/cap_ffmpeg_impl.hpp
//added by xy
#define CODEC_FLAG_GLOBAL_HEADER   (1 << 22)

Pip install numpy

sudo pip install numpy

Face Recognition

/home/xy2/workspace/git/opencv/face-recongnition-20180429

#!/usr/bin/python
#-*- coding: utf-8 -*-

import numpy as np
import cv2

print "============================================="
#print "get camera"
#print "============================================="
#cam = cv2.VideoCapture(0)

#print "--> get width, height"
#print "============================================="
#width = int(cam.get(cv2.CAP_PROP_FRAME_WIDTH) + 0.5)
#height = int(cam.get(cv2.CAP_PROP_FRAME_HEIGHT) + 0.5)
#print width, height

#fourcc = cv2.VideoWriter_fourcc(*'mp4v')
#out = cv2.VideoWriter('output.mp4', fourcc, 20.0, (width, height))

def Getface(img):
        # create classifier
        cvo = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
        cvo.load('/usr/local/share/OpenCV/haarcascades/haarcascade_frontalface_default.xml')
        gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

        # face recognition
        faces = cvo.detectMultiScale(
                gray,
                scaleFactor=1.3,
                minNeighbors=5,
                minSize=(30, 30),
                flags=cv2.CASCADE_SCALE_IMAGE
        )

        # face rect
        for (x, y, w, h) in faces:
                cv2.rectangle(img, (x, y), (x+w, y+h), (0, 255, 0), 2)

        return img

#while(cam.isOpened()):
while(True):
        frame=cv2.imread('demo.jpg')
        faceframe=Getface(frame)
        cv2.imshow('MyCamera', faceframe)

        # quit with keyboard 'q'
        if cv2.waitKey(1) & 0xFF == ord('q'):
                break

print "--> release resource"
print "============================================="
#out.release()
#cam.release()
cv2.destroyAllWindows()

print "--> finished"

作者：七点水Plus
链接：https://www.jianshu.com/p/b50dc83503d3
来源：简书
著作权归作者所有。商业转载请联系作者获得授权，非商业转载请注明出处。