项目作者: Raphpontes

项目描述 :
This repositiry aims to analysys restaurant flow on real time.
高级语言: C
项目地址: git://github.com/Raphpontes/restaurant-flow-analysis.git
创建时间: 2019-10-20T21:13:31Z
项目社区:https://github.com/Raphpontes/restaurant-flow-analysis

开源协议:

下载


Restaurant flow analysis

Authors: Raphael Pontes e Willian Beltrão

This repository aims to analyze restaurant flow on real time, for this purpose, it uses concepts of IoT and Machine Learning . This project was implemented on restaurant university at UNICAMP(University of Campinas - Brazil).

For more details, access to our report in portuguese.

Project Topology

The project topology is showed below:

Arquitetura

There are two relevant parts, it is hall sensor with ESP8266, and camera with raspberry py 3B+.

How to use

For this project, it is necessary to install some libraries for detect people using sensor hall and computer vision.

Sensor hall

For sensor hall, it is necessary to install some libraries on your arduino plataform since this project is based on ESP8266:

1) ESP8266WiF - Library that allows the ESP8266 to connect and communicate via WIFI.

2) PubSubClient - Library that allows to use protocol MQTT and communicate to the server doing simple messages publish/subscribe.

3) ArduinoJson - Library that allows to create files, and in this case messages, on Json Format. This in important library since many servers accept this JsonFormat.

Vision

For vision, the two most important requirements are:

1) Python - Programming language used for the scripts, which organize processing and results.

2) OpenCV - Main library used to work with images for operations such as reading and writing.

References

This project uses other important projects, such as:

1) Yolo Tiny Darknet-nnpack - https://github.com/zxzhaixiang/darknet-nnpack

2) NNPACK - https://github.com/shizukachan/NNPACK

3) OpenCV DNN and MobileNet-SSD - https://github.com/rdeepc/ExploreOpencvDnn