ABSTRACT

Today, assessing the accuracy of a map generated from remotely sensed imagery is a routine component of most mapping projects. However, this was not always the case. By the end of World War II, the use of aerial photography and photo interpretation was a well-established means of learning about the earth’s resources. Maps generated from aerial photos for such uses as agricultural monitoring, forest inventory, geologic exploration, and many others became commonplace. A key component of every photo interpretation project included the necessary ¦eld visits to train the interpreter to recognize the objects of interest on the ground and on the aerial photos. It was generally recognized that the human interpreter drew lines on the photo (i.e., created polygons) around areas that seemed to be distinct from each other (i.e., had more variation between polygons than within a polygon) and then did their best to appropriately label the polygons. Ÿose areas that were di¶cult to label were checked in the ¦eld during another ¦eld visit. Very little thought was given to any

Acronyms and De¦nitions ................................................................................................................ 583 29.1 Introduction ............................................................................................................................ 583 29.2 Assessing Map Accuracy ....................................................................................................... 583 29.3 Positional Map Accuracy Assessment ................................................................................. 585

29.5 Conclusions.............................................................................................................................. 600 References ............................................................................................................................................ 600

type of quantitative evaluation or accuracy assessment of the resulting photo-interpreted maps.