Student research opportunities
Object detection and pose estimation with (RGBD) camera
Project Code: CECS_728
This project is available at the following levels:
Honours, Summer Scholar, Masters
Keywords:
Computer vision, Human computer interface, Bionic eyes, Machine Learning
Supervisor:
Dr Xuming HeOutline:
This project aims to provide a better indoor object recognition and/or pose estimation system using the new Kinect sensor by incorporating both RGB/intensity and depth. The new camera enable us not only to recognize objects in everyday life more effectively, but also to retrieve precise pose information for better human-computer or -object interaction.
Goals of this project
A software demo capable of recognizing indoor objects and providing position and orientation information, which can be used to control the computer or help visually-impaired people.
Requirements/Prerequisites
Basic computer vision and machine learning knowledge.
c/c++ programming experience (familiar with OpenCV/PCL will be a plus).
Student Gain
Knowledge of computer vision techniques with the new RGBD sensor, and potential research opportunities at postgraduate level.
Background Literature
http://www.cs.washington.edu/ai/Mobile_Robotics/rgbd-workshop-2011/
http://www.cs.washington.edu/ai/Mobile_Robotics/projects/3d-object-recognition/
http://www.openni.org/
http://pointclouds.org/
