6.11 Image processing fundamentals for machine learning
You can move an image between representations, apply classical filtering and feature extraction, and feed the result to a classifier.
Before:05. Classical Machine LearningUnlocks:07. Natural Language Processing11. The ECE Bridge — Signals, Edge & Embedded AI
Before convolutional networks there was a craft of filters, edges and hand-designed features, and it has not stopped being useful: it is what remains when data is scarce or compute is tight, and it demystifies what the first convolutional layers learn on their own. This topic moves images between representations, applies classical filtering, and feeds the results to a classifier. It sits just before CNNs so that what they automate is something you have done by hand.
Work through these
Represent an image as an array and move between colour spaces
An image is a numeric array with a channel dimension, and the colour space chosen changes which features stand out. Moving between representations is the first practical skill here.
Apply filtering, edge detection and morphological operations
Classical operations that smooth, sharpen, find boundaries and clean up shapes. They predate learned features and remain the right answer for many well-defined tasks.
Extract classical image features and say when they still beat learned ones
Hand-designed descriptors of texture, shape and keypoints, which still win when data is scarce or the task is narrow. Knowing they exist keeps you from training a network unnecessarily.
Build a feature representation for a small image set and classify it
Building a feature representation for a small image set and classifying it without deep learning. It sets an honest baseline for everything that follows.
Sign in to keep your progress.
Free resources
We haven't checked most of these for screen reader use yet.
Links last checked 29 Aug 2026.
Stuck here?
Ask a mentor. A real person answers, and they can see exactly which topic you're on. Usually within a couple of working days.
Checking your session…
Topics shown in module order.