install-opencv.md

Installing OpenCV

Installing OpenCV using Pip

Pip is the easiest way to install OpenCV.

Dave Jones (creator of the Picamera Python module) and Ben Nuttall (Raspberry Pi community) run piwheels.org, a Python package repository providing pre-compiled binaries packages for the Raspberry Pi.

There are four OpenCV packages that are pip-installable on the PyPI repository:

• opencv-python

  This repository contains just the main modules of the OpenCV library.

• opencv-contrib-python

  This repository contains both the main modules along with the contrib modules.

• opencv-python-headless

  Same as opencv-python but no GUI functionality.

• opencv-contrib-python-headless

  Same as opencv-contrib-python but no GUI functionality.

If you need non-free algorithms you need to compile and install OpenCV from source.

1. Install dependencies

sudo apt-get update && sudo apt-get upgrade

sudo apt-get install \
    libhdf5-dev libhdf5-serial-dev libhdf5-100 \
    libqtgui4 libqtwebkit4 libqt4-test python3-pyqt5 \
    libatlas-base-dev \
    libjasper-dev

2. Create a virtual environment

mkvirtualenv pipcv4 -p python3

3. Install OpenCV

# Activate Python environment
workon pipcv4

# Install numpy
pip install numpy

# Install OpenCV with contrib
pip install opencv-contrib-python

Installing OpenCV from source

1. Install dependencies

Let's update the system:

sudo apt-get update && sudo apt-get upgrade

Install developer tools including CMake:

sudo apt-get install build-essential cmake unzip pkg-config

Install a selection of image and video libraries:

sudo apt-get install libjpeg-dev libpng-dev libtiff-dev \
  libavcodec-dev libavformat-dev libswscale-dev libv4l-dev \
  libxvidcore-dev libx264-dev

Install GTK, a GUI backend (only if you have a Raspberry Pi with Desktop):

sudo apt-get install libgtk-3-dev
sudo apt-get install libcanberra-gtk*

Install two packages for numerical optimizations:

sudo apt-get install libatlas-base-dev gfortran

Install the Python 3 development headers:

sudo apt-get install python3-dev

Create and activate a Python environment:

mkvirtualenv cv4

And finally, install numpy:

pip install numpy

2. Downloading OpenCV files

Copy and paste the following commands:

OPENCV_VERSION="4.1.0"
OPENCV_URL="https://github.com/opencv/opencv/archive/${OPENCV_VERSION}.tar.gz"
OPENCV_CONTRIB_URL="https://github.com/opencv/opencv_contrib/archive/${OPENCV_VERSION}.tar.gz"

cd ${HOME}

curl -L ${OPENCV_URL} -o "opencv.tar.gz"
curl -L ${OPENCV_CONTRIB_URL} -o "opencv_contrib.tar.gz"

tar -xzf opencv.tar.gz
tar -xzf opencv_contrib.tar.gz

mkdir -p opencv-${OPENCV_VERSION}/build
cd opencv-${OPENCV_VERSION}/build

3. Configuring OpenCV installation

# Configure OpenCV via CMake
cmake -D CMAKE_BUILD_TYPE=RELEASE \
    -D CMAKE_INSTALL_PREFIX=/usr/local \
    -D OPENCV_EXTRA_MODULES_PATH="../../opencv_contrib-${OPENCV_VERSION}/modules" \
    -D ENABLE_NEON=ON \
    -D ENABLE_VFPV3=ON \
    -D BUILD_TESTS=OFF \
    -D OPENCV_ENABLE_NONFREE=ON \
    -D INSTALL_PYTHON_EXAMPLES=OFF \
    -D BUILD_EXAMPLES=OFF ..

Check if the interpreter points to the correct Python 3 binary, and if numpy points to the NumPy package installed inside the virtual environment.

• -D OPENCV_ENABLE_NONFREE=ON flag ensures that you'll have access to SIFT/SURF and other non-free algorithms.
• NEON is an optimization architecture extension for ARM processors designed specifically for faster video processing, image processing, speech recognition, and machine learning.
• VFPv3 is the name for ARM's Vector Floating Poing (VFP) version 3.

4. Increase the Raspberry Pi SWAP

Before the compile process, it is recommended to increase the swap space. More swap space enables to compile LibRealSense with all cores of the Raspberry Pi without having problems due to full memory.

Edit and restart the swap service:

sudo sed -i 's/CONF_SWAPSIZE=100/CONF_SWAPSIZE=2048/g' /etc/
dphys-swapfile

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

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

5. Make & make install

make -j4

sudo make -j4 install

sudo ldconfig

6. Reset the the Raspberry Pi SWAP

Remember to go back and reset the swap size:

sudo sed -i 's/CONF_SWAPSIZE=2048/CONF_SWAPSIZE=100/g' /etc/dphys-swapfile

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

We enabled a large swap only to compile LibRealSense. It is not prudent to keep a large swap size for a long time. Increasing swap size is a great way to burn out your SD card. Flash-based storage have limited number of writes you can perform until the card is essentially unable to keep the recorded data.

7. Sym-link OpenCV to your virtual environment

Copy and paste the following commands on your terminal:

# Get the OpenCV library path
opencv_lib="$(ls /usr/local/lib/python3.*/site-packages/cv2/python-3.*/cv2.*.so)"

# Activate cv4 Python environment
workon cv4

# Get the site-packages path of the virtual environment
site_packages="$(python3 -c "import distutils.sysconfig as s; print(s.get_python_lib())")"

# Create a link to OpenCV library
ln -s ${opencv_lib} ${site_packages}

8. Test installation with Python

# Check if the OpenCV library is well configured
python -c "import cv2;  print(cv2.__version__)"
# The output should be 4.1.0

References

[1] Adrian Rosebrock. OpenCV Tutorials, Resources, and Guides. PyImageSearch.

[2] Adrian Rosebrock. Install OpenCV 4 on your Raspberry Pi. PyImageSearch, September 26, 2018.

[3] Adrian Rosebrock. Pip install opencv. PyImageSearch. September 19, 2018.

[4] Ben Nuttall. New opencv builds. Piwheels. September 27, 2018.

[5] Sagi Zeevi. Build a faster OpenCV for Raspberry Pi3. October 30, 2017.

[6] Scott Robinson. Python Virtual Environments: A Primer. Real Python. December 01, 2018.

[7] Shane Pfaffly. How to change Raspberry Pi's Swapfile Size on Raspbian. BitPi.co . February 11, 2015.