Abstract
Panchromatic aerial photography is expected to come into widespread use in the near future because GIS has made it relatively easy to make digital orthophotography using personal computer systems. In this study, unsupervised classification, which is a comparatively simple method of image classification, was derived from the characteristics of histograms of forests, and a quartic regression classification method was devised to distinguish between undisturbed and disturbed forestland that existed after the construction of skidding roads. The error in accuracy of the classification method was 1.18m (2pixels). An evaluation of the classification accuracy by confusion matrix proved that this method was superior to the ISODATA method. Ground vegetation recovered within about 5 years after construction of the skidding roads, and there was no increase in landslides. Regarding impacts to the landscape scenery, it became clear that the period in foliate season was about 10 years until crown closure of the stand around the skidding roads and reforestation.
