wenlong-liu / PyBerryNet

A Python API for BerryNet: A Gateway for Deep Learning On Raspberry Pi

PyBerryNet

A Python API for BerryNet.

Requirements

• Raspberry Pi 3 Model B (Or newer)

• Python libraries:

  • Paho-MQTT
  • BerryNet

Install

To use this library, you need to properly setup Rapsberry Pi 3, install MQTT protocol and configure berrynet services.

Setup Rapsberry Pi 3

Please follow this to set up your Pi.

Configure BerryNet

DT42 group have kindly provided an installment guideline:

$ git clone https://github.com/DT42/BerryNet.git
$ cd BerryNet
$ ./configure

It will take a while to build the programs. Please refer to the GitHub page for further info.

Install paho-mqtt 1.3.0

pip install paho-mqtt

Install pyBerryNet from github

$ git clone https://github.com/wliu2016/PyBerryNet
$ cd PyBerryNet
$ sudo python3 setup.py install

Install pyBerryNet via pip

pip install pyberrynet

Usage

After completing all the installment, the usage will be very simple:

import pyberrynet
berrynet = pyberrynet.run()
results = berrynet.upload('picamera')
berrynet.close()

Roadmap and priorities

Items	Priority	Status
Save result image	*****	Ongoing
Able to draw boxes in the result image	****	Ongoing
Able to change more configuration settings	****	Not started
Formulate documents in readthedoc.com	****	Not started
Fix conflict of backend status within one session	****	Ongoing
Add examples and tutorials	***	Ongoing
Add test suitcase	******	Done
Prepare for pip installment	*****	Done
Testing with RPi3	*****	Done
Add license file	*****	Done

Comments or questions:

Please reply to this issue for problems or comments.