| Abstrakt: |
The paper outlines an analysis of compliance of historical panorama representation using spatial transformation. Although this method deals usually with maps (aerial maps, scanned rasters, historical works), the reliability of this unconventional experimen t may be supported by theoretical principles of cartography and descriptive geometry. The fresco has been spatially transformed using a georeferencing tool, e.g. it was given spatial positions according to the points identified on the drawing as well. A reference coordinate system is usually rel ated to the Earth's surface, but in this case it is not horizontal plane to which are the objects projected. The veduta represents a projection of the town to a general plane (close to the vertical actually) that is supposed to be identical with the image plane of the drawing. Thus the image coordinate system given by the drawing frame was defined and used as the reference for both works of art. A spatial transformation was performed in opensou rce QGIS (QGIS Development Team, 2016). Several transformation methods were tested and a polynomial transformation of 1st order was chosen as optimal. Preliminary transformation proved that the fresco is quite shifted comparing to the drawing. The most significant deformation and the tilt (perspective deformation of original photograph) can be seen on the Castle and under the castle hill. Moreover, there seems to be quite a wide gap between the Castle and the central town. These results were verified transforming the fresco using a smaller number of identical points that were patterned more regularly. This experimen t indicates that there is a gap indeed. Thus the space between the Castle and the central town was narrowed gradually to estimate the optimal distance empirically. Statistical characteristics and comparison with the sketch shows that the gap should be narrowed significantly to keep the distance between the Castle and the central town. However, significant objects (the Castle, towers in the central town, River Danube) match quite well in the horizontal direction, but there is inaccuracy in the vertical placement of the so-called Water Tower and in fact the castle hill seems to be much higher in the fresco than in the sketch. It would be appropriate to verify this using orthogonal photoplan with high resolution. This experiment was not meant to be an artistic or historic validation of the drawing or the fresco. The purpose was to quantify the differences of the artworks. A way to implement exact mathematical methods in historical research was demonstrated. The geospatial approach was forced to abandon the usual point of view and to perceive a kind of geographic data from the position of an artist - landscape painter. [ABSTRACT FROM AUTHOR] |
