Visual Perception and Robotic Manipulation Springer Tracts in Advanced Robotics Chapter 7 System Integration and Experimental Results Geoffrey Taylor Lindsay Kleeman Intelligent Robotics Research Centre (IRRC) Department of Electrical and Computer Systems Engineering Monash University, Australia Overview • • • • • Stereoscopic light stripe scanning Object Modelling and Classification Multicue tracking (edges, texture, colour) Visual servoing Real-world experimental manipulation tasks with an upper-torso humanoid robot Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 2 Motivation • To enable a humanoid robot to perform manipulation tasks in a domestic environment: – A domestic helper for the elderly and disabled • Key challenges: – – – – – Ad hoc tasks with unknown objects Robustness to measurement noise/interference Robustness to calibration errors Interaction to resolve ambiguities Real-time operation Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 3 Architecture Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 4 Light Stripe Scanning Scanned object D B Camera Stripe generator • Triangulation-based depth measurement. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 5 Stereo Stripe Scanner Scanned object X Left image plane xL xR Laser diode θ Left camera L Right image plane Right camera R 2b • Three independent measurements provide redundancy for validation. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 6 Reflections/Cross Talk Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 7 Single Camera Result Single camera scanner Robust stereoscopic scanner Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 8 3D Object Modelling • Want to find objects with minimal prior knowledge. – Use geometric primitives to represent objects • Segment 3D scan based on local surface shape. Surface type classification Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 9 Segmentation • Fit plane, sphere, cylinder and cone to segments. • Merge segments to improve fit of primitives. Raw scan Surface type classification Final segmentation Geometric models Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 10 Object Classification • Scene described by adjacency graph of primitives. • Objects described by known sub-graphs. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 11 Modeling Results • Box, ball and cup: Raw colour/range scan Textured polygonal models Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 12 Multi-Cue Tracking • Individual cues are only robust under limited conditions: – Edges fail in low contrast, distracted by texture – Textures not always available, distracted by reflections – Colour gives only partial pose • Fusion of multiple cues provides robust tracking in unpredictable conditions. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 13 Tracking Framework • 3D Model-based tracking: models modelled from light stripe range data. • Colour (selector), edges and texture (trackers) are measured simultaneously in every frame. • Measurements fused in Extended Kalman filter: – Cues interact with state through measurement models – Individual cues need not recover the complete pose – Extensible to any cues/cameras for which a measurement model exists. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 14 Colour Cues • Filter created from colour histogram in ROI: – Foreground colours promoted in histogram – Background colours supressed in histogram Captured image used to generate filter Output of resulting filter Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 15 Edge Cues Sobel mask directional edges Combine with colour to get silhouette edges Predicted projected edges Fitted edges Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 16 Texture Cues Rendered prediction Feature detector Outlier rejection Matched templates Final matched features Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 17 Tracking Result Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 18 Visual Servoing • Position-based 3D visual servoing (IROS 2004). • Fusion of visual and kinematic measurements. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 19 Visual Servoing • 6D pose of hand estimated using extended Kalman filter with visual and kinematic measurements. • State vector also includes hand-eye transformation and camera model parameters for calibration. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 20 Grasping Task • Grasp a yellow box without prior knowledge of objects in the scene. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 21 Grasping Task Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 22 Pouring Task • Pour the contents of a cup into a bowl. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 23 Pouring Task Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 24 Smell Experiment • Fusion of vision, smell and airflow sensing to locate and grasp a cup containing ethanol. Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 25 Summary • Integration of stereoscopic light stripe sensing, geometric object modelling, multi-cue tracking and visual servoing allows robot to perform ad hoc tasks with unknown objects. • Suggested directions for future research: – – – – Integrate tactile and force sensing Cooperative visual servoing of both arms Interact with objects to learn and refine models Verbal and gestural human-machine interaction Taylor and Kleeman, Visual Perception and Robotic Manipulation, Springer Tracts in Advanced Robotics 26