SCENIC BEAUTY ESTIMATION
OF
FORESTRY IMAGES

Rising public concern for preserving the beauty of natural environments and the need to preserve the aesthetic quality of forests was responsible for the enforcement of various legislative acts to preserve the beauty of forests. In this talk, a collection of algorithms developed to extract features for automatically determining the scenic beauty of a forestry image, and methods employed to relate these features to the scenic beauty estimate will be presented. A system using color, density of trees and entropy achieved an error rate of 36.8%, and a correlation of 0.59 with human judgements.