Installation and media artist François Quévillon was amongst the many practitioners discussing their work at the International Marketplace for Digital Arts presentations during the BIAN two weeks ago in Montreal. Quévillon used his talk to give a brief overview of his experiments with ‘urban imaging’ over the last several years and he showed a range of interesting slitscan-like processed photographs and some more recent work inspired by LiDAR datasets and augmented mapping. Dérive makes use of photogrammetry, geomatic data and 3D modelling to construct point cloud abstractions of buildings, landmarks and cityscapes within a custom ‘browser’ interface.
Dérive capitalizes on it’s underlying coordinate systems by connecting browser and vertex behaviours to environmental data related to the sites of various models. So, a point cloud of a structure in Orléans, France would be inflected by local meteorological and astronomical data and reconfigure itself each time these variables changed. Quévillon’s description of how the system works:
The display and positions of the points and of the wireframe connecting them are determined by the following environmental information as a mean to evoke or simulate them.
Local time: Point size and brightness (relative to sunrise and sunset)
Temperature: Point color
Cloudiness: Point saturation and brightness
Wind: Point displacement reflecting speed and direction
Visibility: Intensity of a depth of field effect and transparency
Humidity: Depth of field focus distance and point sharpness
Precipitation: Random lines drawn from the sky are connected to the ground
This data is coupled with information about atmospheric pressure to yield a generative score for the dynamic animations. Quévillon has implemented Dérive in video loop and installation contexts, with the latter using computer vision to allow viewers to interact with the visualization.
Posted on: 18/05/2012
Posted in: openFrameworks