| Title |
A Study on the Analysis of Building Form Image by Digital Image Processing |
| Keywords |
Image Processing ; Form Analysis ; Correlation coefficient |
| Abstract |
Users' demands on the building form are changing in variety. These demands include reasonable building form, harmonious composition with surroundings and esthetic satisfaction that could be brought by personal tastes and preferences. In addition, models that are introduced from designing process and from various forms tend to lack objective decision making standard. Accordingly it is difficult to find a clear alternative plan and process.
In an effort to solve these problems, the objects of this study are; to propose an analysis model of building form and space by using image process techniques that are on study in the field of artificial intelligence based on acquisition of digital image and to verify the application possibilities and efficiency of such analysis model.
A matrix model for the building form was produced through the proposed analysis model, after which the experiment was carried out. From the form analysis, an influence from linear elements such as edge detection analysis was found. From the building form model analysis, an efficient analysis method proved to be that of light and shadow of building, i.e. binary image process algorism. For these analyses, multiple regression equation that can make quantification of intensity of light and recognition amount extracted from transformed images is used. The transformed images are obtained from independent variable evaluation content standards (edge detection, thinning, binary, block binary) with the matrix model formed from single element transformation as their objects. |