Neural network/Haar cascade trained self-driving RC car using Raspberry Pi
This repository houses all of the software/instructions, as well as schematics, that will allow you to turn an ordinary RC car into a "smart" self-driving car . The car is capable of autonomously detecting the bounds of the road
All of the software in this project (except for a smidgeon of HTML/CSS/JS) is Python 3.6.0.
Note: because this uses the picamera
module, this will likely only run on the rasperry pi.
make run-web-server
##Things you'll need Below is the list of parts that I used to make this project happen.
Note on the Raspberry Pi: I used the Model 2B but the 3 has since come out, either or should work for this project. I would recommend the 3 as you don't need to worry about purchasing an extra WiFi dongle. A kit like this is probably best as it comes with a case, preinstalled SD, etc. For this walkthrough I'll make the assumption that you already have a functioning, WiFi connected Pi.
Note on the RC Car: If you're crafty you really don't need to purchase an RC car. Essentially you'll be "tearing the guts" out of the car down to just the DC motor that controls forwards/reverse, servo that controls left/right, and the chassis. You could build a car with a kit like this for cheaper but I decided not to focus on that (at least for now).
Note on soldering: soldering isn't really necessary for this project, although it's a temporary solution. If you're a noob at soldering like me, you can just strip your wires back, inter-twine the cooper and throw some electrical tape on it.
Note: Fritzing schematic can be found here.
...test scripts to make sure everything works fine