Image Classification

About this course

Design image classification workflows that connect useful map legends, representative training data and defensible accuracy assessment.

This practitioner course is part of the Earth Observation & Remote Sensing school. Work through the curriculum below in a self-paced / instructor-led format, with an estimated study time of 16 h.

Course curriculum

• Module 1 — Classification Design and Legends
• Module 2 — Training Data and Its Biases
• Module 3 — Pixel, Object and Parcel Classification
• Module 4 — Classical and Machine-Learning Classifiers
• Module 5 — Features from Spectra, Texture and Time
• Module 6 — Accuracy Assessment
• Module 7 — Area Estimation with Confidence Intervals
• Module 8 — Map Production and Documentation

Applied learning

• Case Study — The Map That Looked Right
• Lab 1 — Choosing a classifier
• Lab 2 — A land-cover map with a stratified reference sample
• Lab 3 — Area estimates for the Statistics Office
• Capstone — A Validated Land-Cover Product

Who this course is for

GIS / Remote Sensing Professional; Researcher; Government Professional

Prerequisites

NE-RS-202

Course details

• Course code: NE-RS-205
• Level: Practitioner
• Estimated duration: 16 h
• Delivery: Self-Paced / Instructor-Led