Robotic grasping has been a prevailing problem ever since humans began creating robots to execute human-like tasks. The problems are usually due to the involvement of moving parts and sensors. Inaccuracy in sensor data usually leads to unexpected results. Researchers have used a variety of sensors for improving manipulation tasks in robots. This thesis focuses specifically on grasping unknown objects using mobile service robots. An approach using convolutional neural networks to generate grasp points in a scene using RGBD sensor data is proposed. Two neural networks that perform grasp detection in a top down scenario are evaluated, enhanced and compared in a more general scenario. Experiments are performed in a simulated environmentas well as the real world. The results are used to understand how difference in sensor data can affect grasping and enhancements are made to overcome these effects and to optimize the solution. This thesis is an improvement on the works of Douglas Morrison, Peter Corke and Jürgen Leitner in their work Closing the Loop for Robotic Grasping: A Real-time, Generative Grasp Synthesis Approach and Fu-Jen Chu, Ruinian Xu and Patricio A. Vela in their work Real-world Multi-object, Multi-grasp Detection.
Grasping Unknown Objects Using Convolutional Neural Networks
2020-09-30
Hochschulschrift
Elektronische Ressource
Englisch
DDC: | 629 |
Task Understanding for Grasping Unknown Objects
British Library Online Contents | 1993
|