Multimedia  

 

Volume 23 Issue 4/5- Publication Date: 1 April-May 2004
 
Special Issue on the 8th International Symposium on Experimental Robotics (ISER ’02)
 
Vision In and Out of Vehicles: Integrated Driver and Road Scene Monitoring
 
N. Apostoloff Department of Systems Engineering, Research School of Information Sciences and Engineering, The Australian National University, Canberra, ACT, 2611, Australia and A. Zelinsky Department of Systems Engineering, Research School of Information Sciences and Engineering, The Australian National University, Canberra, ACT, 2611, Australia
 

One of the more startling effects of road related accidents is the economic and social burden that they cause. In OECD countries (the 23 leading economically developed countries of the world) over 150,000 people are killed every year (44,000+ in the USA, 38,000+ in Europe and 11,000+ in Japan) at an estimated cost of US$ 500 billion. One way of combating this problem is to develop intelligent vehicles that are self-aware and act to increase the safety of the transportation system. In this paper we present preliminary results of an Intelligent Transport System project that has fused visual lane tracking and driver monitoring technologies in the first step towards closing the loop between vision inside and outside the vehicle. Experimental results of a novel 15 Hz visual lane tracking system will be discussed, focusing on the particle filter and cue fusion technology used. The results from the integration of the lane tracker and the driver monitoring system are presented with an analysis of the driver’s visual behavior in several different scenarios.

 
Multimedia Key
= Video = Data = Code = Image
 
Extension
Type
Description
1
Example One: Overview of the vision systems on TREV. (4.6MB)
2
Example Two: Example faceLAB™sequence showing head pose and eye gaze tracking. (3.9MB)
3
Example Three: Example lane tracking sequence from a high curvature road showing the convergence of particles onto the lane location. (10.3MB)
4
Example Four: The three lane tracking sequences. 1.Ahighway with clear lane markings, shadows and several run-off lanes. 2. A high curvature outer city road showing dramatic lighting changes, a significant shadows and discontinuous lane markings. 3. A highway with poor lane markings and strong shadows. (8.5MB)
5
Example Five: The integrated driver and road scene monitoring system is shown in a 3 minute sequence around a high curvature outer city road. (3.9MB)
6

Example Six: Example dataset from a realworld test as well as the visualization software that integrates the lane tracking results with the head pose and eye gaze vectors from faceLAB™(The face-LAB™Toolbox for MatLAB was kindly supplied by Seeing Machines - which I have extended, with the help of David Liebowitz, to use dynamic patch data). Extract the data and code using your favorite archiving utility and run the “run_viov”
script in a MatLAB shell to visualize the data. The data viewer is a modified version of the threed_browser that comes with FAT v1.0. (0.9MB) ZIP file
7

Example Seven: Example dataset from the highway lane tracking scenario containing the baseline, the respective image set and a transformation framework between the road-centric coordinate system and the image coordinate system. Extract the data and code using your favorite archiving utility and use am_skeleton.m as a template file for your code. (8.1MB) ZIP file
8
Example Eight: Lane tracking sequence along a high curvature outer city road in medium rain. (9.4MB)
 
Return to Contents