Installing OpenCV - FAR-Lab/Developing-and-Designing-Interactive-Devices GitHub Wiki

OpenCV is a great library with computer vision tools and algorithms. You can use OpenCV for things like face detection, object detection, and image manipulation.

For class, we are providing two ways of getting OpenCV onto your IxE:

Burning an updated OS to your SD card
Installing OpenCV on your own

1. Burning a new OS with OpenCV pre-installed (~ 45 minutes)

This is easiest if you are connected to the DeviceFarm network.

Burning a new version of the IxE version of the Raspbian OS is the fastest way to get OpenCV, but it means you will overwrite your entire SD card and start with a fresh system. It will still have all our examples built in, but none of the code you have written. Thus we will backup your code and then reinstall the OS.

Back up your code

Make sure your code is backed up on Github. You can go into each project directory that you are working on and commit and push your code to Github.com. Instructions for using Git with your code can be found here

Inside of a project directory that is git enabled, run

git commit -m "Write your commit message here"
git push

Copy any other files off your Pi to your laptop

If you have other files, like images, sounds, or code that is not backed up on a Git repo, you can copy the files to your laptop. To do this, we have enabled the Samba Filesharing service on your IxE. Samba will allow you to browse the Pi's file system using your laptop native file explorer (Finder on Mac, Explorer on Windows). Here is how to connect to the Samba file share.

First, make sure your laptop is on the same network as your IxE (for example both computers are on DeviceFarm or your own WiFi).

On Mac

Open Finder
Click Go -> Connect to Server in the top toolbar
In Server Address put smb://ixe[00].local replacing [00] with your IxE number.
Put in your username (default is pi) and password to your IxE in the new window and click connect.
Select the homes volume and click Ok
Browse your file system and copy your file using drag and drop or Cmd+C and Cmd+V paste onto your laptop.

On PC

Go to the Windows File Explorer
In the left navigation pane, right-click My Computer
Choose Map Network Drive
In the location put in \\ixe[00]\homes replacing [00] with your IxE number
Uncheck Reconnect at login
Check Use different credentials
Click connect
Put in your IxE username (default: pi) and password into the pop-up. You may need to click Choose another user
Click connect, a new Explorer window should pop up and show your files
Browse your file system and copy your file using drag and drop or Ctrl+C and paste Ctrl+V onto your laptop.

Make sure to do backup everything you want to keep! Once you start burning a new OS there is no way to get your files back.

Download our latest IxE Raspbian OS image onto your laptop

You can download the latest OS here: https://www.dropbox.com/s/pp2lpznt6kd1vl5/IxE_2018-02-22.img.zip?dl=0

Download and install Etcher

Etcher is a nice GUI program that can help you burn the IxE OS image to your SD card. Download the version for your laptop OS.

Shut down your IxE

While logged into your IxE using the Terminal, shut down the IxE using

sudo halt

After the green light stops blinking, pull out the power cable.

Take your SD card and mount on your laptop

Remove the SD card from the Raspberry Pi. Us an SD card adapter and plug it into your computer. If you have an SD card slot, you can use a micro-SD to SD card adapter (available on our cart in the Maker Lab). If you do not have an SD card slot, you can use a micro-SD to USB 3 adapter (there will be a pink adapter on the cart in the Maker Lab -- please do not lose this!).

Burn the IxE disk image using Etcher

Open Etcher
Click Select Image and choose IxE_10FEB2018.img from the location where you download the disk image file
The SD card should be selected automatically, but if not, click Select Drive and choose the SD card. It should be about 15 GB.
Click Flash!

Etcher will then flash the IxE disk image to your SD card. This should take about 25 minutes.

Optional: Setup your WiFi

If you are not on DeviceFarm and want to use your home WiFi or the House WiFi, follow the instructions here to edit the wpa_supplicant.conf file in the boot partition so that you can connect to your own Wifi.

Reboot your IxE

Once the image is flashed/burned to the SD card, you can remove it from your computer and reinsert the card into your Raspberry Pi. Make sure the Raspberry Pi's power cable is unplugged.

Once the card is back in the Raspberry Pi, plug in the power to turn on the Pi.

Since we have reflashed the Pi, it will only have the DeviceFarm WiFi preprogrammed (unless you did the previous step and set it up for your own WiFi). Make sure that your laptop is connected to the same network as your Pi.

Log in and rename your IxE

Since this is the base image, your Pi will be named ixe00. To connect, you can open a terminal and ssh

ssh pi@ixe00

The password is raspberry

Once you are logged in, change the hostname to your original number (written on the SD card we gave you). To do this you will change two files, /etc/hostname and etc/hosts/

Edit /etc/hostname using

sudo nano /etc/hostname

Change ixe00 to ixe[00] where [00] is your number

Save and exit using Ctrl + X, yes

Edit /etc/hosts using

sudo nano /etc/hosts

Change the last line of the file from

127.0.1.1       ixe00

127.0.1.1       ixe[00]

replacing the [00] with your number.

Save and exit using Ctrl + X, yes

Reboot your IxE

For the hostname change to take effect, reboot using

sudo reboot

Reconnect to your IxE

With the hostname changed, you should be able to reconnect to you IxE using your original number. You can ssh using

ssh pi@ixe[00].local

Put your files back on

After you log in, you can reconnect to the Samba file share to put back all the files you copied over to your laptop. You can do this using drag and drop

You can also put your Git repositories back on using

git clone https://github.com/YOUR-REPO-HERE

You will need to do this for every repository you would like to have back on your Pi.

Test OpenCV with NodeJS

You can test that OpenCV works with NodeJS by trying the face-detection.js sample in the node-opencv-test directory.

cd node-opencv-test/
node face-detection.js

If everything is working correctly, then you should see the following output:

Image saved to out.jpg!

You can see the file by typing ls and listing at the files in the directory

ls
face-detection.js  node_modules  out.jpg  package-lock.json

To see the actual image, you can open the file by using the Samba file sharing feature, discussed here or you can copy it to your laptop using scp from another Terminal window on your laptop. The following line will copy the file from your IxE to your Desktop. Remember to change the [00] with you IxE number.

scp pi@ixe[00].local:~/node-opencv-test/out.jpg ~/Desktop

Use `node-opencv` in your own projects

Now that everything is working, you can start using OpenCV in your own NodeJS projects. You can install the library to a project using

npm install opencv --save

To use the library in your javascript code, include the require line at the top of your code file

var cv = require('opencv');

Check out the examples to see how things work and to build your code off the great work of Peter Braden.

2. Installing OpenCV 2.4 on your own (~ 1.5 hours)

This guide will show you how to install a version of OpenCV optimized for your Raspberry Pi onto your IxE.

This guide is based on the tutorial by Adrian Rosebrock at PyImageSearch https://www.pyimagesearch.com/2017/10/09/optimizing-opencv-on-the-raspberry-pi/. He creates easy to follow guides for working with OpenCV. He primarily used Python, which is included on your Pi, but will not be covered in the class.

We will be setting up our IxE to allow NodeJS to use OpenCV. There are two main things we need to install, the general purpose OpenCV library and the node-opencv package.

We will install OpenCV 2.4.13.5, the latest OpenCV 2.4 version at the time this guide was written. OpenCV 2.4 is used because it has better compatibility with NodeJS. Although OpenCV 3 is the newest version, OpenCV 2.4 is still actively maintained.

Prepare your system

1. Make sure you have enough free space on your system using

df -h

You should see that the first line tells you how much space you have in /dev/root

Filesystem      Size  Used Avail Use% Mounted on
/dev/root        15G  4.4G  9.2G  33% /
devtmpfs        434M     0  434M   0% /dev
tmpfs           438M     0  438M   0% /dev/shm
tmpfs           438M   12M  427M   3% /run
tmpfs           5.0M  4.0K  5.0M   1% /run/lock
tmpfs           438M     0  438M   0% /sys/fs/cgroup
/dev/mmcblk0p1   41M   21M   21M  51% /boot
tmpfs            88M     0   88M   0% /run/user/1000

In this case, we have 9.2 GB free (AVAIL) on the system. If you have less than 1 GB then you should clean up your system (see the footnotes of this guide)

2. Install dependencies

In order to install OpenCV, we need to install other packages that OpenCV will use. Run each line of the following commands to install the dependencies. The whole process should take between 5 - 10 minutes.

sudo apt-get update && sudo apt-get upgrade
sudo apt-get install build-essential cmake pkg-config
sudo apt-get install libjpeg-dev libtiff5-dev libjasper-dev libpng12-dev
sudo apt-get install libavcodec-dev libavformat-dev libswscale-dev libv4l-dev
sudo apt-get install libxvidcore-dev libx264-dev
sudo apt-get install libgtk2.0-dev libgtk-3-dev
sudo apt-get install libcanberra-gtk*
sudo apt-get install libatlas-base-dev gfortran
sudo apt-get install python2.7-dev python3-dev
sudo apt-get autoremove

Download OpenCV 2.4

We need to install OpenCV from the source code. The following commands will download the official OpenCV source code to your IxE and unzip the archive.

$ cd ~
$ wget -O opencv.zip https://github.com/Itseez/opencv/archive/2.4.13.5.zip
$ unzip opencv.zip

Compile and install OpenCV with optimizations

1. Create the build directory and setup for compilation

First, go into the OpenCV 2.4 directory and create a build directory where we will compile OpenCV

cd opencv-2.4.13.5/
mkdir build
cd build

Next, compile OpenCV with the following settings. Specifically, enable_NEON=ON and enable_VFPV3=ON. These tell OpenCV to take advantage of some hardware optimizations of the ARM processor that the Raspberry Pi has. You can read more about this from Adrian Rosebrocks guide.

Copy all the commands below into your terminal and then hit enter/return.

cmake -D CMAKE_BUILD_TYPE=RELEASE \
    -D CMAKE_INSTALL_PREFIX=/usr/local \
    -D ENABLE_NEON=ON \
    -D ENABLE_VFPV3=ON \
    -D BUILD_TESTS=OFF \
    -D INSTALL_PYTHON_EXAMPLES=OFF \
    -D BUILD_EXAMPLES=OFF ..

2. Increase the swap space for faster compiling

Increase the swap space to make installing faster and allow you to use all four cores on the Raspberry Pi. We will change this only for completing the install of OpenCV and then change it back since this can shorten the life of your SD card.

Edit the dphys-swapfile using sudo

sudo nano /etc/dphys-swapfile

Change the line CONF_SWAPSIZE=100 to CONF_SWAPSIZE=1024

Save and exit the file using Ctrl+X, yes

Restart the swap service

sudo /etc/init.d/dphys-swapfile stop
sudo /etc/init.d/dphys-swapfile start

3. Compile and install OpenCV (takes ~40 minutes)

Now we will compile OpenCV. This will take a long time since the Raspberry Pi processors is not that powerful. Even when using four cores, it will take about 40 minutes to compile everything.

Run make with the four core compile option -j4

make -j4

Wait until the OpenCV has full compiled. It will reach 100% and then should go back to the command prompt and allow you to type commands. Do not power off your Pi while you are compiling.

After OpenCV is compiled, install and configure using

sudo make install
sudo ldconfig

4. Decrease your swap space

Now that OpenCV is installed, reduce the swap space back to the original value of 100

sudo nano /etc/dphys-swapfile

Change the line CONF_SWAPSIZE=1024 to CONF_SWAPSIZE=100

Save and exit the file using Ctrl+X, yes

Restart the swap service

sudo /etc/init.d/dphys-swapfile stop
sudo /etc/init.d/dphys-swapfile start

Install node-opencv into a project

Now that you have OpenCV installed, let's create a new NodeJS project and install the node-opencv package, created by Peter Braden. This will allows us to use OpenCV with NodeJS.

1. Create a new project

First, create a new project directory where we can test the library

cd ~
mkdir node-opencv-test
cd node-opencv-test

2. Install node-opencv

Run

npm install opencv

This will take about 5 minutes since it will compile a number of packages.

3. Run the test

To check and make sure everything is working, we will run the face-detection example. This code will detect a face in an image then put a red circle around it. We will run the example with an image of the Mona Lisa included in the examples directory of the opencv package.

Create a new file called face-detection.js

nano face-detection.js

Then copy the following code into the editor

var cv = require('opencv');

cv.readImage("node_modules/opencv/examples/files/mona.png", function(err, im){
  im.detectObject(cv.FACE_CASCADE, {}, function(err, faces){
    for (var i=0;i<faces.length; i++){
      var x = faces[i]
      im.ellipse(x.x + x.width/2, x.y + x.height/2, x.width/2, x.height/2);
    }
    im.save('out.jpg');
    console.log('Image saved to out.jpg!')
  });
})

Save and exit the file using Ctrl+X, yes

Run the code using NodeJS

node face-detection.js

If everything is working correctly, then you should see the following output:

Image saved to out.jpg!

You can see the file by typing ls and listing at the files in the directory

ls
face-detection.js  node_modules  out.jpg  package-lock.json

scp pi@ixe[00].local:~/node-opencv-test/out.jpg ~/Desktop

Use `node-opencv` in your own projects

Now that everything is working, you can start using OpenCV in your own NodeJS projects. You can install the library to a project using

npm install opencv --save

To use the library in your javascript code, include the require line at the top of your code file

var cv = require('opencv');

Check out the examples to see how things work and to build your code off the great work of Peter Braden.

Installing OpenCV - FAR-Lab/Developing-and-Designing-Interactive-Devices GitHub Wiki

1. Burning a new OS with OpenCV pre-installed (~ 45 minutes)

Back up your code

Copy any other files off your Pi to your laptop

On Mac

On PC

Download our latest IxE Raspbian OS image onto your laptop

Download and install Etcher

Shut down your IxE

Take your SD card and mount on your laptop

Burn the IxE disk image using Etcher

Optional: Setup your WiFi

Reboot your IxE

Log in and rename your IxE

Reboot your IxE

Reconnect to your IxE

Put your files back on

Test OpenCV with NodeJS

Use node-opencv in your own projects

2. Installing OpenCV 2.4 on your own (~ 1.5 hours)

Prepare your system

1. Make sure you have enough free space on your system using

2. Install dependencies

Download OpenCV 2.4

Compile and install OpenCV with optimizations

1. Create the build directory and setup for compilation

2. Increase the swap space for faster compiling

3. Compile and install OpenCV (takes ~40 minutes)

4. Decrease your swap space

Install node-opencv into a project

1. Create a new project

2. Install node-opencv

3. Run the test

Use node-opencv in your own projects

Use `node-opencv` in your own projects

Use `node-opencv` in your own projects