3D-reconstruction from vision is used for a number of applications:
- To monitor tolerances and detect deviations in subtractive or additive manufacturing
- To continuously inspect the 3D-shape of cutting tools, in order to avoid production or quality loss because of defects
- To improve the localization of objects of interest for pick & place operations, for manipulations or quality inspection, and for navigation purposes
In all of those applications, current 3D-sensing technology is insufficient because it is
- Does not combine the information in a single consistent world model
The goal of the project is to improve accuracy, resolution and speed of 3D-imaging techniques, and let them provide consistent 3D-information. The project will not develop new hardware, but make use of state-of-the-practice technology in 3D, 2D, and 1D-imagery. Yet, we will develop the methodology to optimally choose and configure combinations of different sensors: which and how many sensors, what hardware specifications, where to place, what low-level configurations, etc.
Active3D is an ICON-project (Interdisciplinair Coöperatief Onderzoek – Interdisciplinary Cooperative Research). We are looking for companies to participate in the project, as they will help us define the industrial use cases for the project.