Research

Share    

Using stereo vision on a fast moving unmanned ground vehicle

Objective

To develop visual SLAM, obstacle detection and tracking as well as road feature detection system for a high-speed unmanned ground vehicle using real-time stereo vision.


Description

The focus of this project is on developing computationally efficient algorithms that support obstacle detection and terrain classification using stereo vision techniques. The development of long-range advanced stereo vision techniques, as compared to other techniques that depend on active vision i.e. lidar, enables low-cost implementation of obstacle detection and collision avoidance for autonomous vehicles. The stereovision system aims to achieve both distant object detection up to 50m and robust localisation at a maximum speed of 50 km/h.


Result

                                                                             Lane detection

      

                                                       Detection and segmentation of obstacle up to 50m.


                                                          Real-time loop closure scenario at 50 km/h


Principal Investigators


Associate Prof. Wang Han (NTU)

Telephone: 6790 4506
Office: S2-B2b-49
Email: hw@ntu.edu.sg
Mr. Paul Tan (STE)

Telephone: 6660 1052
Office: S1-B4a-03
Email: paultan@stengg.com