UAV Panoramic Images for Waterfront Landscape Analysis and TopographicDB Texturing

Abstract

The integration of landscape into planning policies requires an articulated territorial approach supported by multi-dimensional and multi-temporal Spatial Data Infrastructure. It requests to improve SDI able to generate innovative typologies of representation with respect to ‘different points of view’. In fact, many landscape representations rely nowadays on models integrated by aerial images with weak possibilities to reproduce views similar to the human eye, especially the ones at a human height. A methodological approach of rediscovering and correlating to new current maps the long tradition of historical human eye perspectives, water front view maps and itinerary maps, has been here experimented. The paper investigates the potential of panoramic view reconstruction and simulation from images acquired by RC/UAV and by multisensor ground platforms. The main aim is to generate multiple views models able to simulate real scenarios at the height of the observer, at ground level and lower altitudes. In particular the use of panoramic images for TopographicDB texturing is discussed showing their potential in improving value content decoding of built environment and 3D city models, supporting planners with more realistic instruments in the Environmental Impact Assessment (EIA). At the same time, such representations can improve knowledge transfer to citizens, toward an acknowledged tourism and volunteered image data acquisition through I-devices applications. On the other side, limits and critical elements that need to be overcome by the research of new algorithms and new acquisition instruments are shown.