ABSTRACT

Critical steps in analyzing any map generated from remotely sensed data are an error analysis and an accuracy assessment. The purpose of quantitative accuracy assessment is to identify and measure map errors. There are four major reasons for performing an accuracy assessment:

To understand the errors in the map (so they can be corrected)• To provide an overall assessment of the reliability or usefulness of the map [1]• To use the map in a decision-making process• Because the contract demands it•

Assessing the accuracy of maps generated from digital remotely sensed data has a short history, beginning in the mid-1970s. Before this time, maps derived from analog (film) remotely sensed data (i.e., photo interpretation) were rarely evaluated using any kind of quantitative accuracy assessment. Field checking was performed as an important component of the interpretation process, but no overall map accuracy or other quantitative measure of quality was generally undertaken. It was only after photo interpretation began to be used as the reference data to compare digital remotely sensed data, did anyone begin to seriously consider how accurate the photo interpretation really was. In fact, the photo interpretation was often assumed to be 100% correct, resulting in some very poor digital image accuracies-in reality, some of these poor accuracies were due to errors in the photo interpretation and not due to the digital image analysis.