See images like a computer
- Images as Arrays
- IP 02 · Digital Image Fundamentals
- IP 03 · Intensity Transformations and Spatial Filtering
- IP 04 · Filtering in the Frequency Domain
Free · English and 繁體中文 · runnable code in every chapter
A course in image processing, machine learning and computer vision, with the language, optimization and information theory around them. Every idea starts in plain words, then the math, then code you can run.
How every chapter works
Each section opens with an everyday explanation, before any symbol appears.
Definitions and equations, with every symbol explained right after it.
Short, tested Python whose printed output matches the text.
Exercises with hints close every chapter, and interactive experiments let you change a value and watch what happens.
Try it now
Histogram equalization spreads the narrow band of gray levels in a low-contrast image across the whole range. IP 03 explains why it works and where it goes too far.
Open IP 03 →

Learning paths
What an agent is, how tool calling and the context window actually work, where agents fail, and how to choose the smallest system that finishes your task.
How to read model evidence (what a benchmark measures, contamination, uncertainty, repeated runs), what 'open' means at each licensing layer, and how to run your own evaluation.
How image generators work, which capabilities matter for a given deliverable, what benchmarks and metrics can and cannot tell you, and how to run a fair acceptance trial.