8000 GitHub - iiyyll01lin/Dr._Fram: Auto Defect Detection with Autonomous Driving, Object Detection & Discrimination, and a Web App.
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

Auto Defect Detection with Autonomous Driving, Object Detection & Discrimination, and a Web App.

Notifications You must be signed in to change notification settings

iiyyll01lin/Dr._Fram

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

README

Dr._Farm

Intro

Dr Farm is an Auto Defect Detection Project with Autonomous Driving, Object Detection & Discrimination, and a Web App.

Dr Farm is a project consisting of 3 parts: autonomous driving, object detection and web application.

Pi_Donkeycar & Pi_Camera

The autonomous driving part is inherited from Donkeycar (https://github.com/autorope/donkeycar), which is trained to specialize in fram tracks; the object detection part is trained using YOLOv4-tiny model, and the web app part uses Node.js to connect to two Raspberry Pis on the car, this is why we have two directories, Pi_Donkeycar and Pi_Camera, to represent the separate deployment.

Ubuntu_train

The Ubuntu_train directory contains how to train the self-diving models by Keras. The model is quantized by TensorFlow Lite and pruned in some scale to deploy on Raspberry Pi successfully. I left some models in Ubuntu_train/mycar/ I trained for your reference. BTW, the mypilot_0308_base.h5 model is the most stable for my track, FYI.

Web_App

The Web_App directory contains all the Servlets, Node.js, and views to support the connection between two Raspberry Pis and the Website.

About

Auto Defect Detection with Autonomous Driving, Object Detection & Discrimination, and a Web App.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published
0