1. Accuracy Assessment Accuracy Assessment Error Matrix Sampling Method
Geography KHU Jinmu Choi
Accuracy Assessment
Error Matrix
Sampling Method

2. Accuracy Assessment
The thematic information must be accurate because important decisions are made throughout the world using the information.
Unfortunately, the thematic information contains error. Scientists who create remote sensing–derived thematic information should recognize the sources of the error, minimize it as much as possible, and inform the user how much confidence he or she should have in the thematic information.
Remote sensing–derived thematic maps should normally be subjected to a thorough accuracy assessment before being used in scientific investigations and policy decisions.

3. Steps Required to Perform Accuracy Assessment
(source: Jensen, 2011)

4. Error Matrix
To locate ground reference test pixels (or polygons if the classification is based on human visual interpretation) in the study area.
Not used to train the classification algorithm
(source: Jensen, 2011)

5. Sample Size
Fitzpatrick-Lins (1981) suggests that the sample size N to be used to assess the accuracy of a land-use classification map be determined from the formula for the binomial probability theory:
where p is the expected percent accuracy of the entire map, q = 100 – p, E is the allowable error, and Z = 2 from the standard normal deviate of 1.96 for the 95% two-sided confidence level.
(source: Jensen, 2011)
With expected map accuracies of 85% and an acceptable error of 5%
acceptable error of 10%

6. Sampling Design 1. random sampling, 2. systematic sampling,
3. stratified random sampling,
4. stratified systematic unaligned sampling, and
5. cluster sampling.
(source: Jensen, 2011)

7. Descriptive Statistics
The overall accuracy of the classification map is determined by dividing the total correct pixels (sum of the major diagonal) by the total number of pixels in the error matrix (N ).
The probability of a reference pixel being correctly classified is a measure of omission error. This statistic is the producer’s accuracy because the producer of the classification is interested in how well a certain area can be classified.
The probability that a pixel classified on the map actually represents that category on the ground is a measure of commission error. This measure, called the user’s accuracy because the user of the classification is interested in how much the classified area is correct.

8. Kappa Analysis Khat Coefficient of Agreement:
Kappa analysis yields a statistic, , which is an estimate of Kappa. It is a measure of agreement or accuracy between the remote sensing–derived classification map and the reference data as indicated by a) the major diagonal, and b) the chance agreement, which is indicated by the row and column totals referred to as marginals.
(source: Jensen, 2011)

9. Error matrix of the classification map derived from hyperspectral data of the Mixed Waste Management Facility on the Savannah
River Site.
(source: Jensen, 2011)

10. Next Lab: Exercise 12: Unsupervised Classification
Project Presentation
Final Exam
Source:
Jensen and Jensen, 2011, Introductory Digital Image Processing, 4th ed, Prentice Hall.