Image Processing in JavaScript
summary
JavaScript implementations of concrete image techniques: adaptive thresholding for binarization, three segmentation approaches including stack vs queue tradeoffs (one example shows a max stack of 119) and a flood fill demo, a face localization pipeline using Haar-like features and cascade classifiers (real classifiers have thousands of features and use AdaBoost/OpenCV trainers), and a shipped tool, the Dimensions Chrome extension, that measures pixels by taking scrolling screenshots, converting to grayscale and flood‑filling — it works well on flat designs but struggles with gradients.
AI-written summary · about
chapters (7)
- 0:00REJECTE
- 3:23Segmentation
- 7:21Face Localization
- 13:38Dimensions
- 14:27Open Source
- 14:57Demo Time
- 15:52The End
related talks
32mMagic of Computer Vision Unraveled
36mMaking a Robot Eye with JavaScript
26mSelf-Portraits in JavaScript
32mSelf-Portraits in JavaScript
26mHow to Cheat at Life with JS, CV, and AR
01h 13mChallenges and Trade-offs for In-Browser Computer Vision Applications
29mImage Processing in the Browser
01h 42mTeaching Robots to See With Javascript