https://github.com/ricklove/blaze-car

Blazing Fast Self Driving RC Car - 200FPS Image Analysis on Raspberry Pi
https://github.com/ricklove/blaze-car

Last synced: about 1 month ago
JSON representation

Blazing Fast Self Driving RC Car - 200FPS Image Analysis on Raspberry Pi

• Host: GitHub
• URL: https://github.com/ricklove/blaze-car
• Owner: ricklove
• Created: 2019-07-26T20:14:30.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2019-08-06T10:47:49.000Z (over 5 years ago)
• Last Synced: 2024-11-01T18:37:47.000Z (about 1 month ago)
• Language: C
• Size: 171 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-blazingly-fast - blaze-car - Blazing Fast Self Driving RC Car - 200FPS Image Analysis on Raspberry Pi (C)

README

        
# Blaze Car

Blazing Fast Self Driving RC Car 

200FPS Image Analysis on Raspberry Pi

## Summary

Full Convolutional Neural Networks are simply too much processing for a Raspberry Pi 3B - especially at 200FPS. 

A more custom approach is needed running on a constrained environment.

## Goals

- Support speeds of at least 30mph

- Adapt to any visual conditions:

  - Indoor support with glare from indoor lighting

  - Outdoor support with glare from sun and shade over the course

- Learn extreme driving modes like sliding turns and drifting 

- Map course solely from visual input

## Modules

- Pixel Online Learning Module

  - Learn Pixel Colors Online (during runtime) to adjust to lighting conditions

  - To ensure stability, use only verified results - strategically using a more expensive algorithm like edge detection to confirm learning results

# To Build and Run

  - ./buildme

  - ./run