10. Image Classification

Objectives

This learning object introduces the contents of unit 10, Image Classification.

This is tenth unit in the module “Remote Sensing for Earth Observation”. The objective of this unit is to introduce the concept and practice of image classification, which is the central operation in transforming remotely sensed data into information about the surface of the Earth in the familiar form of land cover mapping. There are five learning objects:

  1. Introduction to Unit 10
  2. What is classification?
  3. Types of classification
  4. Unsupervised classification
  5. Supervised classification
  6. Object based classification

‘What is classification?’ introduces the process of image classification and particularly the concept of a multidimensional feature space onto which individual pixel values and land cover types can be mapped.

‘Types of classification’ introduces the two principal approaches to image classification, namely unsupervised and supervised classification. It briefly considers their advantages and disadvantages, before they are addressed in more detail individually by the following objects:

‘Unsupervised classification’ explains the general operation of unsupervised image classification, with a brief over view of the k-means and ISODATA algorithms.

‘Supervised classification’ explains the process of supervised image classification and covers the stages involved: user definition of land cover classes; training/site selection; generation of statistical parameters; classification. Parallelpiped, minimum distance to means and maximum likelihood classification approaches are reviewed.

Object based classification gives a brief overview of the classifcation approaches which use image ‘objects’ rather than ‘single’ pixels to delineate land cover/land use.

There are practical and reflective activities designed to reinforce students’ learning associated with each of the objects. The practical activities will be perfomed using ENVI software and will provide you with knowledge/skills needed to work on Assignment 2.

Comments are closed.