This research project is divided into three stages, where each stage lasts for a year. Next we present the activities to be done for each stage.
Stage 1: Analysis and design of modules for a visual system
- Definition of the vision sub-modules to be used for each module (e.g., region of interest detection).
- Definition of the methodology to be used to integrate those sub-modules into a specific module.
- Algorithms design. An analysis will be done to determine what kind of approach(es) are suitable for the vision problem.
- Development and implementation of prototypes for each defined sub-module, where preliminary results can be obtained.
- Comparison of the prototypes with state of the art techniques.
Stage 2: Implementation and testing in a humanoid robot
- Analysis and design for the oculo-motor system control and other robot motor systems.
- Visual sub-modules integration into their corresponding sub-module.
- Prototype implementation.
- Comparative analysis in terms of eficiency and performance of each module with respect to similar state of the art techniques.
- Implementation of the oculo-motor system control and other motor systems.
Stage 3: Module integration and system evaluation
- Interface design between modules and motor systems.
- Integration of the whole system.
- Comparative analysis in terms of efficiency and performance between current hierarchical visual systems and our proposed system.
- Discussion and analysis of results.
Coordinación de Módulos de Control Guiados Visualmente en un Marco de
Toma de Decisiones para Robots Humanoides
Proyecto CONACyT No. 177041
Last updated:19 January 2014