Classification Basics

Purpose

Outline how image classification converts spectral and spatial measurements into thematic map categories.

Outline

• Classes, labels, and the mapping question
• Features from bands, indices, texture, terrain, and time
• Supervised and unsupervised classification concepts
• Training data quality, sampling design, and class balance
• Accuracy assessment, confusion matrices, and uncertainty

Later Examples

• Separating water, vegetation, bare ground, and built surfaces
• Comparing training samples across classes
• Reading a confusion matrix for map accuracy