Every day the quantity of vehicles on the roads around the world is increasing. This growth combined with the negligence of drivers and some external factors such as road and weather conditions result on a huge rise on the accidents and fatalities statistics. Since the beginning of the 21 century, many research groups and automotive companies are developing and adapting technologies that can be embed on the vehicles in order to reduce these dramatic numbers. One interesting example of these technologies is the detection and classification of mobile obstacles (vehicles, people, etc.) in urban environments. This work presents the development of algorithms for identification, classification, tracking, and prediction of moving obstacles,in special pedestrians. We used laser sensor data for monitoring the environment that surrounds our test vehicle (a modified passenger car). Based on these data we performed a computational treatment in order to classify all detected obstacles into two main classes: static and mobile obstacles. Then, the algorithm tracks and predicts the mobile obstacles positions for a certain time window by applying a Kalman Filter and a simply linear model for obstacle velocities. Even if the mobile obstacle is out of the sensor range (or occluded by other obstacles), the Kalman Filter used can predict its estimated position and trajectory for the time window. Another benefit of using a simple and more generic model is the fact that we are dealing with obstacles that may have different dynamic characteristics (e.g.: cars, motorcycles, bicycles, pedestrians, etc.). Based on the prediction of the obstacle positions, the vehicle navigator (an embedded navigation algorithm) can generate the best path taking into account all detected and hidden obstacles.
[1]
Hans-Hellmut Nagel,et al.
Model-Based Object Tracking in Traffic Scenes
,
1992,
ECCV.
[2]
Roland Siegwart,et al.
Introduction to Autonomous Mobile Robots
,
2004
.
[3]
Charles E. Thorpe,et al.
Simultaneous localization and mapping with detection and tracking of moving objects
,
2002,
Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No.02CH37292).
[4]
Cristiano Premebida.
Segmentation and Geometric Primitives Extraction from 2D Laser Range Data for Mobile Robot Applications
,
2005
.
[5]
Y. Bar-Shalom.
Tracking and data association
,
1988
.
[6]
Markus Kohler,et al.
Using the Kalman Filter to track Human Interactive Motion - Modelling and Initialization of the Kalm
,
1997
.
[7]
Geovany de Araújo Borges,et al.
Line Extraction in 2D Range Images for Mobile Robotics
,
2004,
J. Intell. Robotic Syst..