lakshanthad/retail-store-items-detection-nodered

Get started with AI vision at the edge with no coding experience at all!

Build and Deploy a Retail Store Items Detection System using Node-RED on IoT Edge device

Node-RED is a low-code programming language for event driven applications. It is a programming tool for wiring together hardware devices, APIs and online services in new and interesting ways. It provides a browser-based flow editor that makes it easy to wire together flows using the wide range of nodes in the palette that can be deployed to its runtime in a single-click. JavaScript functions can be created within the editor using a rich text editor. It comes with a built-in library that allows you to save useful functions, templates or flows for re-use.

Hardware Components

• Seeed Studio reComputer J1010 with NVIDIA Jetson device
• USB Camera Module
• Mini Wireless USB Adapter
• Wireless USB Desktop Keyboard/Mouse

Software Components

• JetPack 4.6.1
• NodeJS(for local development)
• Docker
• Docker Compose

Getting Started

1. Wire up the Hardware System

Step 1 - Install

Clone this GitHub repo and run the installer

git clone https://github.com/Seeed-Studio/node-red-contrib-ml
cd node-red-contrib-ml && sudo ./docker-ubuntu.sh

Step 2 - Configure

Open a web browser, type jetson_device_ip_address:1880 on the search box, drag and drop blocks and connect them as follows

Step 3 - Deploy

Press DEPLOY to see it in action!

Application

retail-store-items-detection.mp4