Camera object classification | Blueprints gallery
Blueprints / Camera object classification

Camera object classification

Classifying objects using Edge Impulse and Particle

Intermediatev1.0.0CameraIdentificationAIML
Introduction
Supported devices
Hardware and devices
Cloud services and integrations
Project description

Introduction

Integrate Edge Impulse’s ML with Particle devices to classify objects from a connected camera in real time.

Supported devices

Hardware and supplies

Cloud services and integrations

  • edge-impulseEdge Impulse

Project description:

This project demonstrates how to use Edge Impulse to classify objects in front of a camera connected to Particle devices like the Photon 2 or M-SoM. With Edge Impulse’s machine learning capabilities, this project can recognize and classify objects in real time, ideal for smart home systems, security applications, and more.

Key Features:

  • Edge Impulse Integration: Leverage Edge Impulse’s ML models to detect objects.
  • Device Compatibility: Compatible with Particle Photon 2, Boron, and M-SoM devices.
  • Customizable Model: Pre-trained to detect lamps, plants, and pianos, but can be easily retrained to identify other objects.

Steps include:

  • Hardware Setup: Connect a uCAM-III camera module to your Particle device.
  • Data Collection: Use Edge Impulse to gather and label training images.
  • Model Training: Train a custom model on Edge Impulse for object detection.
  • Deployment: Deploy the model on Particle and start recognizing objects in real-time.
Fork this project to your GitHub account Fork and edit