2019:Vision: Difference between revisions

From 1511Wookiee
Jump to navigationJump to search
(Created page with "<div class="mw-parser-output"><div class="mw-parser-output"><div class="mw-parser-output"> = Hardware  = *Raspberry Pi Model 3 *3D printed case w/ a 40mm fan *Power su...")
 
m (fix code tags)
 
(2 intermediate revisions by the same user not shown)
Line 1: Line 1:
<div class="mw-parser-output"><div class="mw-parser-output"><div class="mw-parser-output">
<div class="mw-parser-output"><div class="mw-parser-output"><div class="mw-parser-output">
<div class="mw-parser-output">
= Hardware&nbsp; =
= Hardware&nbsp; =


*Raspberry Pi Model 3  
*Raspberry Pi Model 3  
*3D printed case w/ a 40mm fan  
*3D printed case w/ a 40mm fan  
*Power supply: 12V->5V regulator (Current Logic FRC-360-CPR)  
*Power supply: 12V->5V regulator (CTRE VRM)  
</div>


= Software =
= Software =
Line 12: Line 14:
*Python3&nbsp;  
*Python3&nbsp;  


<div class="mw-parser-output">
<div class="mw-parser-output">
= Pi Setup =
= Pi Setup =


&nbsp;
&nbsp;


<code>Get .zip of custom image from network</code>
<code><font face="monospace">Generate customized image using&nbsp;</font>[https://github.com/frc1511/pi-gen https://github.com/frc1511/pi-gen]&nbsp;(Linux required)</code>


<code>&nbsp;</code>
<code>&nbsp;</code>
Line 62: Line 66:
<code>Go to ``7 Advanced Options`` and select ``A1 Expand Filesystem``, then reboot.</code>
<code>Go to ``7 Advanced Options`` and select ``A1 Expand Filesystem``, then reboot.</code>


<br/> &nbsp;
<br/> &nbsp; </div>
</div>


= Software Compilation =
= Software Compilation =

Latest revision as of 18:09, 21 April 2019

Hardware 

  • Raspberry Pi Model 3
  • 3D printed case w/ a 40mm fan
  • Power supply: 12V->5V regulator (CTRE VRM)

Software

  • Customized Raspbian Stretch
  • OpenCV with FFMpeg and GStreamer
  • Python3 

Pi Setup

 

Generate customized image using https://github.com/frc1511/pi-gen (Linux required)

 

Use Etcher (https://www.balena.io/etcher/), Rufus (https://rufus.ie/), or any other program that can write an image to a SD card

 

Get a microSD card of at least 8GB in size but no more than 32GB - if using already-setup image use a 32GB card

 

Use the imaging tool to write the image to the SD card

 

Use a USB power supply > 2A

 

Connect a HDMI monitor and keyboard to the Pi and put the SD card in the slot

 

Power up the Pi and wait for it to boot

 

After booting, use the credentials below to log in

    Username: pi

    Pasword: raspberry

 

Run ``sudo raspi-config``, select ``4 Localization Options`` then ``I1 Change Locale``. Scroll to en_US and select all 3 locales with Space, followed by Enter to generate the locales.

 

Go back to ``4 Localization Options`` and select ``I3 Change keyboard layout``.

 

Go to ``7 Advanced Options`` and select ``A1 Expand Filesystem``, then reboot.


 

Software Compilation

Double line break indicates a new line, single line break is continuing the previous line

 

Start with customized pi image

 

sudo apt update && sudo apt upgrade

 

sudo apt install build-essential cmake pkg-config

 

sudo apt install libjpeg-dev libtiff5-dev libjasper-dev libpng12-dev

 

sudo apt install libavcodec-dev libavformat-dev libswscale-dev libv4l-dev

 

sudo apt install libxvidcore-dev libx264-dev

 

sudo apt install libgtk2.0-dev

 

sudo apt install libatlas-base-dev gfortran

 

sudo apt install python2.7-dev python3-dev

 

cd ~

 

wget -O opencv.zip https://github.com/opencv/opencv/archive/4.0.1.zip #latest version as of writing

 

unzip opencv.zip

 

wget -O opencv_contrib.zip https://githib.com/opencv/opencv_contrib/archive/4.0.1.zip #latest version as of writing

 

unzip opencv_contrib.zip

 

wget https://bootstrap.pypa.io/get-pip.py

 

sudo python get-pip.py #if using python2

 

sudo python3 get-pip.py #if using python3

 

pip install numpy
 

Go here if you used the setup script

 

sudo nano /etc/dphys-swapfile and set CONF_SWAPSIZE to 1000. Run ``sudo dphys-swapfile setup``, then reboot with ``sudo reboot``.

 

cd ~/opencv-4.0.1/ #latest version as of writing

 

mkdir build

 

cd build

 

cmake -D CMAKE_BUILD_TYPE=RELEASE \

-D CMAKE_INSTALL_PREFIX=/usr/local \

-D INSTALL_PYTHON_EXAMPLES=ON \

-D OPENCV_EXTRA_MODULES_PATH=~/opencv_contrib-4.0.1/modules \

-D BUILD_EXAMPLES=ON \

-D ENABLE_NEON=ON\

-D ENABLE_VFPV3=ON \

-D ENABLE_FFMPEG=ON \

-D ENABLE_GSTREAMER=ON \

-D PYTHON_DEFAULT_EXECUTABLE=$(which python3) .. #use $(which python2) if using python2

 

Make sure that the section in CMake output includes valid paths for Interpreter, Libraries, numpy, packages path for the Python version used, as well as ‘YES’ for both GStreamer and FFMpeg

 

Make sure that the Pi has appropriate cooling before compiling. It WILL overheat otherwise.

 

DO NOT DO ANYTHING WITH THE PI WHILE COMPILING AS THERE IS A VERY HIGH RISK OF CRASHING

 

make -j4 #compile using all 4 cores, takes a very long time to do (~4 hours)

 

sudo make install

 

sudo ldconfig

 

sudo nano /etc/dphys-swapfile and set CONF_SWAPSIZE back to 100. Run ``sudo dphys-swapfile setup``, then reboot with ``sudo reboot``.

 

rm ~/opencv-4.0.1

rm ~/opencv_contrib-4.0.1