DYNAMIC STRATEGY
FOR BUILDING 3-D
PHOTOREALISTIC MODELS ¹

Henrik Haggrén
Helsinki University of Technology
Institute of Photogrammetry and Remote Sensing

Date:

Funding period: 1.1.1999 - 31.12.1999

Amount of funding applied for: FIM 313,000

Number of person-years: 1.5

Abstract:

In this project, we develop a dynamic strategy for building 3-D photorealistic models of objects and space from raw data acquired either with a multi-camera or single-camera based 3-D video digitizing equipment. A photorealistic model consists of a geometrical part, which contains information about the 3-D shape of the surface, and of a radiometrical part, which tells the tone of the surface given by colors or gray levels.

The dynamic approach is based on the idea that we should take advantage of the vast amount of data the video cameras are able to produce as a function of time. By proper data processing techniques, we intend to improve our 3-D video digitizing system so that the shape of a photorealistic surface can be measured with an accuracy better than 1:10,000 relative to single scene dimensions and the model will have colors or gray levels that are independent of lighting conditions. The dynamic approach will also be applied to guide the measuring into areas where prescribed error tolerances have not yet been achieved.

During the third year of this project, we will develop a self-calibration method to improve the accuracy of the geometry and estimate the precision of the radiometric model. The results of the project can be utilized in industrial applications where accuracy aspects are important.