Intermediate · module I7 · Sublevel Advanced
Computer Vision in Practice
By the end: Process images with OpenCV, explain convolution, and train and deploy a simple image classifier
- Covers
- Working with images in Python, from pixels to trained image models
- Contents
- 3 chapters, 14 topics
- Time
- About 3 hours
- Level
- Intermediate
- Your progress
- 0 / 14
This module builds on Neural Networks and Deep Learning Essentials. We suggest looking at it first, but the choice is yours.
Chapter 1 Images as Data
- 1.110 min Read Activity
Pixels, Channels and Colour Spaces
Grayscale, RGB and HSV as arrays of numbers
Code: inspect an image's numbers
- 1.212 min Read Activity
Introduction to OpenCV
Reading, displaying, resizing, cropping and saving images
Code: build a photo editor
- 1.312 min Read Activity
- 1.410 min Read Activity
Chapter 2 Vision Tasks
- 2.115 min Read Activity
Image Classification
Training a classifier and using a pre-trained CNN
Code: classify everyday objects
- 2.210 min Read Activity
Data Augmentation for Images
Flips, shifts, rotations and noise that teach a model to ignore unimportant changes
Code: shifted digits with and without augmentation
- 2.310 min Read Activity
Object Detection Concepts
Bounding boxes, IoU and how detectors work
Interactive: draw boxes and compute IoU
IoU = overlap / union
- 2.410 min Read Activity
- 2.58 min Read Activity
- 2.610 min Read Activity
Optical Character Recognition
Turning images of text into characters by segmenting and recognising them
Code: read a line of handwritten digits