Multi-visualization of cities

Advances in 3D data acquisition techniques (laser scanning, photography, etc.) has led to a sharp increase in the quantity of available 3D geolocated data. More and more cities provide scanned data on open access platforms. To ensure the intercompatibility of different data sources, standards have been developed for exchange protocols and file formats. Moreover, thanks to new web standards and the increase in processing power of personal devices, it is now possible to integrate rich content, such as 3D applications, directly into a web page. These two elements make it possible to share and exploit 3D city data in a web browser. Several proposals have been done by our team in the last years. Works on nD visualizations have been already described before.

In the work of Jeremy Gaillard, we explored various ways of organizing this data in generic structures to allow the progressive transmission of high volumes of 3D data. Taking into account the multi-scale nature of the city is a key element in the design of these structures. Adapting the visualization of the data to the user is another important objective of my thesis. Because of the high number of uses of 3D city models, the user’s needs vary greatly: some specific areas are of higher interest, and data has to be represented in a certain way… We explored different methods to satisfy these needs, either by prioritizing some data over others during the loading stage or by generating personalized scenes based on a set of preferences defined by the user.

Online demonstrations