education

Education


Round Corner
Department of Computer and Information Science

Oppgaveforslag

GPU implementation of registration of Point Clouds from a RGB-D camera mounted in a robot arm

SINTEF Ocean is leading a research project, called iProcess (www.iprocessproject.com ) aiming to develop novel concepts for flexible robot based automation in the food processing industry. An interesting area of research in WP3 (http://iprocessproject.com/wp-3-flexible-processing-automation/) consists on the vision guided and machine learning based robot manipulation of compliant food objects. Given the 3D images and point clouds from a RGB-D camera mounted on a robot arm, the aim is to develop a grasping concept that finds a suitable grasping pose for the gripper tool.

In this project assignment, the student will work with a GPU implementation of an incremental registration of subsequent point clouds acquired from the Intel RealSense camera from different views with respect to the object. The aim is to perform the registration of point clouds, when the camera (attached to the robot arm) is moved, in real time.
Within a potential follow up master thesis, the objective is to identify the inaccurate or missing parts within the mesh and do a real time registration for a moving object on a conveyer belt.

The project assignment is as following:
• Acquire point clouds with Intel RealSense of objects such as pottery and selected fruits by moving the robot arm.
• Develop appropriate stitching/registration algorithms for the collected point clouds. A good starting point can be Iterative Closest Point (ICP) algorithm.
• Perform a GPU implementation of the algorithm so that the stitching/registration is done in real time.

A state-of-the-art workstation with the necessary software and an appropriate GPU card will be made available for the student. It is preferable to work in C++ or Python making use and building upon existing libraries such as the Point Cloud Library and the VTK toolkit.

Prerequisites:
• Excellent programming skills in C++ and Python
• Knowledge and interest in computer vision, graphics, and machine learning
Co-supervisors from SINTEF Ocean: Dr. Ekrem Misimi – Senior Scientist (ekrem.misimi@sintef.no) and Dr. Christian Schellewald (christian.schellewald@sintef.no ).


The work is available as Autumn project and may continue to Master Thesis and, depending on the quality of the work, SINTEF Ocean may facilitate the publication in a conference or journal paper.

If interested, please send email to Ekrem or Christian by May 7th, 2017.

 

Faglærer

Theoharis Theoharis Theoharis Theoharis
Professor
422 IT-bygget
735 91447 
 
NTNU logo