Scenic evaluation of landscape for urban design purposes using GIS and ANN

The gravitation of population and industry to limited urban-purpose land gives rise to many urban problems such as housing and increase of land prices. The development of mountainous district (MD) which has more available areas than flatlands has an important meaning. The development must be environmentally friendly. In this study, we studied objective landscape estimation and prediction method using geo-spatial information system (GIS) and artificial neural network (ANN) to suggest a desirable MD development model. We compared the landscape preferences of ANN which used only the landscape element value extracted from normalization function with observer groups and verified that the landscape preference of ANN on each landscape is equal to the observer groups. As a result of this, the landscape of the large-scale housing complex developed on MD can be evaluated and predicted using just the values of landscape elements in ANN.