When trying to create machine vision applications, OpenMV is an interesting platform with its goal to become the “Arduino of machine vision.” It combines firmware based on MicroPython with easy-to-use machine vision software libraries, a custom Integrated Development Environment (IDE), and specialized camera hardware. In this tutorial, we use OpenMV and an Arduino Pro Nicla Vision camera board to detect objects like soda cans.

OpenMV Project Overview

In this article, I’ll use the Arduino Pro Nicla Vision camera board (Figure 1) with the OpenMV firmware for object detection. You can, however, run this project on any OpenMV camera hardware. Object detection is a special type of image classification. While an image classification model simply tells you the class of object in the image, an object detection model also indicates the location of the recognized object. This even works for multiple objects in one image. This process is more computationally intensive than simple image...