Syllabus

Photo credit: Dare

Description

Computer vision for manipulators:

The course will start with a short review on motion planning using ROS and Moveit to refresh the participants knowledge on topics covered during the previous year. We will then expand on this and cover how to generate and use a hand-eye calibration. This will be followed by a review of available 3D computer vision technologies and their application. Next strategies for 3D reconstruction using a robot manipulator and depth camera will be reviewed and put into practise. Subsequently the course will delve into capture processing, focussing on point cloud registration, masking and segmentation. Lastly we will cover image segmentation using AI models.

Learning Objectives

The course aim to give students an overview of the state of the art of 3D computer vision for manipulators

Capturing 3D data
- generating hand eye calibration
- utilising hand eye calibration
- capture strategies
Processing capture data
- registration
- segmentation
- marker detection