Raspbuggy : Computer Vision support

Warning message

You must login to view this form.

The Raspbuggy (http://cmcrobotics.github.io/raspbuggy/) is a Wifi-enabled programmable RC car running on top of a Raspberry Pi and a variety of hardware (thanks to Drivar, its Python hardware abstraction layer).

The Raspbuggy Platform is used to teach children and non-programmers about simple logic (loops, tests, variables) and be able to observe the result of the program for real.

 The Raspbuggy programming environment is based on Google Blockly, a drag and drop programming environment targetting complete programming beginners.

Unfortunately, there is currently no way for the Raspbuggy to support a video camera (such as the Pi Camera) and implement OpenCV (or SimpleCV) to add support for image recognitioni.

 

The idea is to implement a couple of new Blockly blocks that add support for SimpleCV to the Raspbuggy.

Category: 
Education
Contacts: 

Brice Copy (brice.copy@cern.ch) - CERN Mobile 165682

Skills Being Sought: 

OpenCV / SimpleCV

Javascript

Python a plus

Weekend Goals: 

Add support for a few new Computer Vision blocks :

* Acquire a picture

  • Keep picture as a template (example : Take picture of a yellow tennis ball)
  •  Compare picture and template - extract data :
    • Is the template present in the picture ?
    • Where in the picture is the template located (e.g. pixel location)
    • How big is the tempate in the picture (e.g. size of the template in pixels)

These simple block allow to add support for image recognition and allow to implement interesting programs :

* Race around a circuit, stop if you see a red light

* Follow the tennis ball

etc...

 

Prerequisites: 
Javascript programming, Python, OpenCV or SimpleCV
Express your Interest to work on this Idea !!

You are here