Skip to content
Repo for Great Uni Hack Manchester 2018
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
assets
src
LICENSE
README.md

README.md

Great Uni Hack 18

Repo for Great Uni Hack Manchester 2018 MLH Hackathon

10-11 November 2018

We made a real-time object-tracking setup utilising the following hardware:

  • Raspberry Pi Model 3B
  • PiCam v2.0
  • Arduino for TTL-to-PWM conversion
  • servos
  • and (most importantly) a LASER!

The objective of our hack is to create a demo of a real-time object tracking system, which can be deployed on existing CCTV hardware in order to track wristbands/markers etc in the public.

The target audience is organisations wishing to track and be aware of vulnerable people in crowds. (The laser is just a cool add-on to make you go wow ;D ).

demo

We won the hardware track and then went on to beat all other tracks to win the grand prize!

laser

Code

Final code is available in src

Code is shown to work on Raspbian Jessie and Stretch.

Software Prerequisites

  • Python 2 (cannot run on Python 3 ;( )

The following Python libs:

  • collections
  • numpy
  • argparse
  • imutils
  • cv2
  • math
  • time (Pi specific)
  • serial (Pi specific)

Known issues and workarounds

On old Raspbian Jessie versions, the specific module needed for cv2.VideoCapture() to access the camera stream does not load on bootup. We certainly spent a few good hours through the night troubleshooting this ;D .

Issue can be prevented by downloading (or building) then adding the following module in /etc/modules:

bcm2835-v4l2

You can’t perform that action at this time.