Connecting Raspberry Pi Pico with a camera to Edge Impulse

I am a beginner working on a project where I would like to use image classification on a RPi Pico, so I made an Edge Impulse project that identifies a deodorant can to start off with. However, I cannot find any documentation on how to process the camera data from the Pico to work with the model. I managed to to build the code on a Raspberry Pi 3B+ (I can also use a Raspberry Pi 4 if that would work better) and get a video stream from the camera on my computer using this tutorial: https://docs.arducam.com/Arduino-SPI-camera/Legacy-SPI-camera/Pico/Camera-Module/SPI-Camera/
But I can’t find any documentation on how to use this code with Edge Impulse. I understand I have to use the data forwarder to take sensor data, but I still do no know how to use image data (The tutorial uses accelerometer data: https://docs.edgeimpulse.com/docs/tools/edge-impulse-cli/cli-data-forwarder). I am a beginner, so my knowledge of using this camera and Edge Impulse isn’t great, so I would greatly appreciate any help.

Project ID: 548248

Environment:

  • Platform: Raspberry Pi Pico, Raspberry Pi 3B+/Raspberry Pi 4
  • Build Environment Details: Raspberry Pi OS
  • OS Version: Raspberry Pi OS 64-Bit
  • Custom Blocks / Impulse Configuration: Images Processing Block and Transfer Learning (Images) Learning Block
    Logs/Attachments:
    image

Hi @stephenswa

The easiest way for starting off is to take the images and upload them to your project, there are many ways to do this programatically too though e.g. REST API that may be easy to implement on the PI : https://docs.edgeimpulse.com/reference/edge-impulse-api/user/upload_photo

Best

Eoin