ISSN:
1432-1009
Keywords:
Forest cover types
;
Small scale aerial photography
;
Vegetation reconstruction
;
Computer-aided classification
;
Computer-aided mapping
Source:
Springer Online Journal Archives 1860-2000
Topics:
Energy, Environment Protection, Nuclear Power Engineering
Notes:
Abstract The US National Park Service must map forest cover types over extensive areas in order to fulfill its goal of maintaining or reconstructing presettlement vegetation within national parks and monuments. Furthermore, such cover type maps must be updated on a regular basis to document vegetation changes. Computer-aided classification of small scale aerial photography is a promising technique for generating forest cover type maps efficiently and inexpensively. In this study, seven cover types were classified with an overall accuracy of 62 percent from a reproduction of a 1∶120,000 color infrared transparency of a conifer-hardwood forest. The results were encouraging, given the degraded quality of the photograph and the fact that features were not centered, as well as the lack of information on lens vignetting characteristics to make corrections. Suggestions are made for resolving these problems in future research and applications. In addition, it is hypothesized that the overall accuracy is artificially low because the computer-aided classification more accurately portrayed the intermixing of cover types than the hand-drawn maps to which it was compared.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1007/BF01876887
