Landscape aesthetic quality assessment of forest lands: an application of machine learning approach
Forests, with natural factors, provide visual aesthetic features, as a social ecosystem service for human beings. Nowadays, forest managers are looking for decision support system tools which evaluate the aesthetic quality of forest landscapes, particularly in the development of human ecosystem serv...
Saved in:
Published in | Soft computing (Berlin, Germany) Vol. 27; no. 10; pp. 6671 - 6686 |
---|---|
Main Authors | , , |
Format | Journal Article |
Language | English |
Published |
Berlin/Heidelberg
Springer Berlin Heidelberg
01.05.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Forests, with natural factors, provide visual aesthetic features, as a social ecosystem service for human beings. Nowadays, forest managers are looking for decision support system tools which evaluate the aesthetic quality of forest landscapes, particularly in the development of human ecosystem services. In our methodology, the aesthetic quality of environment is assessed with a human-perception-based approach to apply three machine learning techniques (support vector machine (SVM), radial basis function neural network (RBFNN) and multilayer perceptron (MLP)) for the aesthetic quality simulation of forest areas. To perform this method, the landscape attributes (13 features) were defined in 72 Hyrcanian broad leaves forest landscapes. The landscapes aesthetic quality model was designed to determine the visual qualities by machine learning techniques. Considering the results, MLP model was detected as the most practical, reliable and accurate model for evaluation of landscape quality in broad leaves forest areas. Comparing to RBFNN (R
2
= 0.809), and SVM (R
2
= 0.829), MLP (R
2
= 0.878) model represents the most reliable results of R
2
in the test data set. The number of species, tree density,
Alnus subcordata
, canopy density, altitude and
Carpinus betulus
in forest areas were detected as the main influential factors of the MLP model. On the other hand, the designed graphical user interface tool finds the most scenic landscapes for peoples who are looking for relaxation and recreation services in the nature. The forested lands planted according to the ecological techniques could be designed with resulted aesthetics criteria. |
---|---|
ISSN: | 1432-7643 1433-7479 |
DOI: | 10.1007/s00500-022-07642-3 |