From the 1980 Machine Processing of Remotely Sensed Data Symposium


Procedure 1 approaches to developing land cover classifications were compared with the Multicluster Blocks process on a 15,000 hectare forested area in southwestern Colorado. Results showed that P-1 (using the clustering processor in an unseeded, iterative mode) performed as well as the Multicluster Blocks approach on the rugged study area. The average accuracies of classification for the best P-1 method and the McB approach were 77.8 and 75.3 percent respectively; overall accuracies were 88.3 and 87.4 percent respectively.

These results may interest the forestry community or any resource discipline which has available to it ground-checked or photointerpreted point (or plot) information. Procedure 1 can use this information directly to output a land cover classification with little analyst interaction.

Date of this Version