Using a vast archive of fragments of scanned architectural photographs, new architectural images are created as multilayered collages of these fragments. The images, created as complex reconstructions of fragments of real images, introduce a third, abstract point of view to the original views of architect and photographer .
The source images were initially analysed and categorised according to their vanishing points and to the shapes contained within them. Based on this analysis, the images were digitally sliced into many pieces which retain the information of their positions relative to the vanishing-point and lines of perspective in the original image. The fragments thus form a large pool of pieces, ready to be applied to new perspectives and shapes within new images.
In order to compose a collage, a perspective-grid is defined and a covering of matching image segments is applied. The segments are never altered to match the frame but instead, appropriate ones are chosen from the mass of possible pieces. By defining additional keywords which describe the content of the original photographs, the selection of segments used for the final composition can be influenced. Thus a contextual layer is added through the semantic linking with the source material.
The recompositions mix and match the views and perspectives of both the architect and the photographer with a third, newly chosen frame. The resulting fine-art prints are completely unique each time.
FELD developed two independent software applications for this project – one for determining vanishing points in high resolution images and one for compositing and rendering of the final collages. The vanishing point finder app analyses images according to their perspective and determines the present vanishing points in a semi automatic process and slices the picture into smaller fragments.
The compositor app is used for creating a perspective grid to which fitting segments should be found from the archive of image fragments. The program allows the easy manipulation of options and the including of image content describing keywords.