ubuntu16.04 configuration opencv face recognition

---

1. GCC 4.4.x or later
2. CMake 2.6 or later
3. Git
4. GTK + 2.x or higher, including headers (libgtk2.0-dev)
5. pkg-config
6. Python
7. Numpy
8. ffmpeg or libav development packages: libavcodec-dev, libavformat-dev, libswscale-dev

---

Installation Environment

 sudo apt-get install build-essential
 sudo apt-get install cmake git libgtk2.0-dev pkg-config libavcodec-dev libavformat-dev libswscale-dev

sudo pip install numpy

These are the foreplay, next to business.

In the file you want to install opencv folder, enter

git clone https://github.com/opencv/opencv.git 

We installed opencv source code from git

We enter opencv

Next steps so divided:

1. Create a temporary folder, and enter
2. cmake our opencv
3. installation

mkdir release   #创建一个临时文件夹

cd release  #进入它

cmake -D CMAKE_BUILD_TYPE=RELEASE -D CMAKE_INSTALL_PREFIX=/usr/local .. #cmake我们的opencv

make

sudo make install#安装

There step in cmake -D CMAKE_BUILD_TYPE=RELEASE -D CMAKE_INSTALL_PREFIX=/usr/local .., we may have something to download for it, this time we can be manually downloaded from.
Here a link to http://blog.csdn.net/yiyuehuan/article/details/52951574 .

This installed the opencv environment, but if we need to use it in python interface, we also need to download something

---

Interface with python opencv

pip install numpy 
pip install Matplotlib
pip install opencv-python

We need to download the three modules on it.

When we downloaded, you can use the opencv cv2 module for python programmed.

Let's make a simple face recognition, I saw http://blog.csdn.net/Marksinoberg/article/details/52443214 this blog to learn.
But some things have not fit now, and I am here to update it.

# coding:utf-8

import cv2

# 待检测的图片路径
img_path = '2.jpg'
Classifier_path = '/media/asahi/数据/所有下载/opencv-3.3.1/data/haarcascades/haarcascade_frontalface_default.xml'
#这个路径表示，我们下载的opencv中自带的haarcascade_frontalface_default特征集，用的时候只需
#把/media/asahi/数据/所有下载/opencv-3.3.1/，换成你自己的路径就行了


face_cascade = cv2.CascadeClassifier(Classifier_path)

# 读取图片

image = cv2.imread(img_path)

gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)


# 探测图片中的人脸

faces = face_cascade.detectMultiScale(

    gray,

    scaleFactor=1.15,

    minNeighbors=5,

    minSize=(5, 5),

    flags=cv2.CASCADE_SCALE_IMAGE   #因为现在opencv也升级成了3，所以cv2.cv也不存在了

)


for (x, y, w, h) in faces:
    cv2.rectangle(image,(x,y),(x+w,y+w),(0,255,0),2)
    # cv2.circle(image, ((x + x + w) / 2,
    #                    (y + y + h) / 2),
    #            w / 2,
    #            (0, 255, 0),
    #            float(2))

cv2.imshow("Find Faces!", image)

cv2.waitKey(0)

There is also a pit, Ubuntu bloggers here are also equipped with ROS system, if run directly to the error.
such as

Traceback (most recent call last):
  File "/media/asahi/办公/work/py/learning_opencv/learning_01.py", line 10, in <module>
    import cv2
ImportError: /opt/ros/kinetic/lib/python2.7/dist-packages/cv2.so: undefined symbol: PyCObject_Type

Checked, ros should be engaged in something. I have no better way can it /.bashrc the last link to the ros statement to the Notes.

Then restart the terminal, you can run.