Home /Research /Object detection using RGBD data for interactive robotic manipulation
MANIPULATION

Object detection using RGBD data for interactive robotic manipulation

Jun-Sik Kim

Year
2014
Citations
2

Abstract

We discuss about practical consideration of 3D object detection and pose estimation using an RGBD sensor for interactive robotic manipulation. Template matching based object detection with an RGBD data works good, but it requires further pose refinement for a robot to manipulate it. The pose data can be used to augment a virtual object in the scene so that a user can manipulate it intuitively. In addition, an automatic database collection system is introduced for easier use of this system.

Keywords

Computer scienceArtificial intelligenceComputer visionObject (grammar)Object detectionPoseRobotMatching (statistics)Virtual imagePattern recognition (psychology)

Related papers

Browse all MANIPULATION papers