Angela MANCUSO / Andrea PASQUALI
(University of Florence, Department of Architecture, Italy)
Keywords: Photogrammetry, Survey, Comparison, Processing, Web Resources
The aim of our research is to analyze three-dimensional models obtained by processing different survey data, to compare the specific workflows in terms of data gathering speed, performance and processing efficiency, errors and their mitigation methods in final digital models.
The results will be achieved by a comparison between 3D digital models, generated by photogrammetry based on specific set of pictures, but the firsts from a Direct Survey Campaign (DSC) and the second from the Web: the variable of the research is the difference of the data underlying the processes.
The DSC is executed in situ, in a short time, performed by a single operator with a tripod, so the resulting images are uniform for coloring and lighting.
The Web Archive Survey Campaign (WASC) consists in a selection of web images, with the experience of the test developed in October 2011 by the University of Washington, called “Building Rome in a Day”. The result is characterized by many differences of the shots collected: various photographic equipments, light conditions, exposure, web compression alterations, but also many points of view.
After a careful verification of data usability, the two groups of images are processed in dedicated softwares; then the calculation is finalized in a set of 3D-models with texture mapping.
The final surface mesh will be subjected to comparative tests between the other models.
The results obtained will contain the conclusions about the degree of reliability of WASC in front of the complete and exhaustive DSC methods, evaluating perceptual appearance, by image rendering processes, as well as through the reliability of the geometric return on errors observation.
All the case studies will be aimed to Cultural Heritage subjects, analyzed in their specific contests and capable to give a clear reference scenario of these important solutions for quick and well working digital survey.