AI & machine vision for robots
Give your build eyes — classical vision to TinyML on-device.
Created by Lerato Dlamini, Tshega Mampshika
What you'll learn
- Load and process camera images with OpenCV
- Calibrate a camera and correct lens distortion
- Track an object and send the position to a robot
- Train a simple image model and test its errors
Skills you'll gain
- OpenCV
- Computer vision
- Camera calibration
- Robot control
- Image labelling
Part of these pathways
Course content
1 section · 6 lectures · 14 total hours · ≈ 2 MB to download
Module 1
- How robots see — pixels, thresholds, and blobs Locked until you enrol 2 h · < 1 MB
- Classical vision — line and shape detection Locked until you enrol 2 h · < 1 MB
- Datasets — collecting and labelling your own Locked until you enrol 2 h · < 1 MB
- Training a TinyML classifier Locked until you enrol 2 h · < 1 MB
- Deploying on-device — no cloud required Locked until you enrol 2 h · < 1 MB
- Capstone — the vision-guided sorter Locked until you enrol 4 h · < 1 MB
Requirements
- Python experience
- A Raspberry Pi or a laptop with a webcam
Description
Six weeks on perception: how robots see, classify, and react.
We begin with classical vision — thresholds, blobs, and line detection — then train and deploy TinyML models that run on-device, no cloud required. Learners build a vision-guided sorter as the capstone, echoing the bottle-top sorter that took one of our juniors to the SAGE World Cup.
Comfort with basic coding is expected; the Foundation programme or equivalent is the right preparation.
Who this course is for
- Learners in the age group: Ages 15+
- Intermediate-level learners
Instructors
Lerato Dlamini
Senior software engineer and AI trainer
- 9 courses
- 193 learners
Lerato leads the AiX software and AI courses. Lerato has built systems for public-sector clients and teaches the testing and code review habits that the courses depend on, with a focus on code that the next person can maintain.
Tshega Mampshika
Co-founder and CTO, AlgoAtWork
- 14 courses
- 257 learners
Tshega leads technology at AlgoAtWork, the company behind AiX Academy. Tshega leads the robotics, electronics and hardware courses and reviews the design of each pathway.
4.0 course rating · 7 ratings
-
Vusi Ngcobo
A solid intermediate course. I would take it after the robotics bootcamp.
-
Keabetswe Mokgosi
The first two modules lean on theory. It becomes practical after that.
-
Sizwe Mthembu
The camera calibration lesson is long, but it is essential.
-
Shanaaz Abrahams
The model training lab needs a GPU or patience with the Colab limits.
-
Ntombi Khoza
I wish the dataset labelling section covered more tools.
-
Naledi Maseko
OpenCV basics were explained well, and the robot tracking project worked first time on my Raspberry Pi.
-
Tumelo Matlala
Good link between vision and robot control.
Frequently asked questions
Do I need a GPU?
No. The model training lab uses small models, and a free Colab session is enough.
Which robot does the course use?
Any wheeled robot with a microcontroller. The examples use the AiX robot car.