Attempt to introduce machine learning software assuming landscape education

Application to buildings on university campuses as an example

Abstract

In recent years, citizen participation-type landscape creation and awareness-raising activities are gradually accumulating. In particular, the importance of landscape guidelines is increasing not only from the viewpoint of preserving historical landscapes but also from the viewpoint of sharing the history of land use. Based on this background, this paper presents an analysis framework for the process from the extraction of landscape elements to the creation of landscape plans. Specifically, based on a method called machine learning, it is an attempt to apply the framework to landscape education through image search. In particular, this study was verified using machine learning software that has become popular in recent years. As a result, although there are limitations to app development, the image evaluation framework originally envisioned was constructed. In this way, the effectiveness of landscape education was confirmed to some extent. In the future, as image analysis software and application development software become widespread for the general public, this issue is expected to be improved. This time, the framework was applied only to buildings of a university campus, but after this, I would like to consider applying it to the landscape of a village.