This paper presents a night time vehicle detection system performing automatic headlamp control in the frame work of driving assistance. In such application dealing with sensor, processing and actuator, we focus on image processing techniques developed in this project. From our embedded camera, image processing enables to detect vehicles ahead and estimates their positions in order to increase driver visibility by adjusting headlamps. We review algorithms (segmentation, classification, tracking and position estimation) in detail and present results comparing driver dazzling between static headlamps and intelligent headlamps. Our system detection range is above 600m for headlamps and about 400m for tail lamps which is sufficient to avoid glaring of other road users. Classification performances are above 97% of true positive rate evaluated on a validation database (frame by frame detection). The final vehicle detection is guaranteed at 100% of recognition by attributing a minimum confidence accumulated over successive fames. By way of conclusion, we introduce perspective of advanced lighting automation.
[1]
D. Fernandez,et al.
Night time vehicle detection for driving assistance lightbeam controller
,
2008,
2008 IEEE Intelligent Vehicles Symposium.
[2]
R. E. Kalman,et al.
A New Approach to Linear Filtering and Prediction Problems
,
2002
.
[3]
Bernhard E. Boser,et al.
A training algorithm for optimal margin classifiers
,
1992,
COLT '92.
[4]
Joseph S. Stam.
Automatic Vehicle High-Beam Headlamp Control System
,
2001
.
[5]
Fuhui Long,et al.
Feature selection based on mutual information criteria of max-dependency, max-relevance, and min-redundancy
,
2003,
IEEE Transactions on Pattern Analysis and Machine Intelligence.
[6]
Anil K. Jain,et al.
Feature Selection: Evaluation, Application, and Small Sample Performance
,
1997,
IEEE Trans. Pattern Anal. Mach. Intell..
[7]
Michael J. Flannagan,et al.
High-beam headlamp usage on unlighted rural roadways
,
2004
.