Elbrechter, Christof: Towards Anthropomorphic Robotic Paper Manipulation. 2020
Inhalt
- Introduction
- Rich Research Challenge of Paper Manipulation
- A New Image Processing Library
- Requirements
- Alternative Computer Vision Libraries
- OpenCV
- Intel IPP
- Halcon
- Matlab Image Processing Toolbox
- Less Common Libraries
- Point Cloud Library (PCL)
- Comparison
- The Image Component Library (ICL)
- Important Tools for this Work
- Easy to Use Core Functionality
- Grabber Framework for Dynamic Image Source Selection
- 2D and 3D Visualization
- Marker Detection Toolbox
- Soft and Rigid-Body Physics Module
- Discussion and Next Steps
- Picking up Paper
- Related Work
- Perception
- Modeling
- Evaluation
- Fiducial Marker Detection Accuracy
- Qualitative Comparison of Modeling Performance
- Quantitative Evaluation of the Mean Modeling Error
- Distance Preservation Error
- Conclusion
- Robot Control
- Discussion
- Bending and Folding
- Related Work
- Perception
- Modeling
- Evaluation
- Robot Control
- Updated Vision and Robot Setup
- Registration of Reference Objects on the Robot Server
- Closed Loop Feedback Controllers
- Folding Paper With the Robot
- Discussion
- Advanced Aspects
- A Generalized Paper Model
- Kinect-based Paper Detection
- A Kinect-based Prototype for Tracking Paper
- Strengths, Weaknesses and Heuristical Improvements
- Folding the Paper in Half
- Discussion
- Supplementing Point-clouds with 2D-SURF-Features
- Extending the ICP-Pipeline by SURF-feature Detection
- Qualitative Evaluation of Human Folding Sequences
- Tracking Folding of Common Textured Paper
- Discussion
- Automatic Fold Detection and Optimization
- Robotic Manipulation of Paper from a System Perspective
- Conclusion
- Bibliography