Bin-Picking (eBook)

Acknowledgments 7
Contents 8
List of Figures 10
Abstract 13
1 Introduction---Automation and the Need for Pose Estimation 14
2 Bin-Picking---5 Decades of Research 16
2.1 The Early Years: Basic Developments 16
2.2 Modern Bin-Picking Approaches 21
2.3 Yet Another Bin-Picking-Approach? 22
2.3.1 Revisiting Robotic Bin-Picking---Problems to Be Solved 23
2.3.2 Contributions and Organization of This Work 24
3 3D Point Cloud Based Pose Estimation 26
3.1 Generic Pose Estimation Using 3D Point Clouds 27
3.1.1 3D Point Cloud Based Pose Estimation 27
3.1.2 3D Edge Based Pose Estimation 33
3.2 Bin-Picking Application---Collision Avoidance and Grasp Planning 36
3.2.1 Efficient 3D Collision Avoidance 36
3.2.2 Depth Image Based Collision Measurement 38
3.3 Experimental 3D Point Cloud Based Pose Estimation 41
3.3.1 Simulation 42
3.3.2 Real World Scenario 43
3.4 Discussion 50
4 Depth Map Based Pose Estimation 51
4.1 Gripper Pose Estimation 52
4.1.1 Fast Gripper Pose Hypotheses Generation 53
4.1.2 Hypothesis Evaluation and Gripper Pose Estimation 54
4.2 Modifications and Enhancements 57
4.2.1 Pitch and Yaw Angles of the Pick Pose 57
4.3 Bin-Picking Application---Grasp Pose Estimation 58
4.3.1 Vision Based Grasp Pose Estimation 59
4.3.2 Force/Torque/Acceleration Based Grasp Pose Estimation 60
4.4 Experimental Depth Map Based Bin-Picking 64
4.4.1 Hardware 64
4.4.2 Grasping Unknown Objects 65
4.4.3 Bin-Picking 66
4.5 Discussion 68
5 Normal Map Based Pose Estimation 69
5.1 The Normal Map 70
5.2 Generic Pose Estimation Using Normal Maps 70
5.2.1 Orientation Estimation 71
5.2.2 Accurate Monocular Translation Estimation 89
5.3 Bin-Picking Application---Collision Avoidance 96
5.4 Experimental Normal Map Based Grasping 96
5.4.1 Simulation 97
5.4.2 Real World Scenario 103
5.5 Discussion 106
6 Summary and Conclusion 108
Appendix A Data Acquisition 111
Own Publications and References 122