Laserscanner Innovations for Detection of Obstacles and Road
This paper presents innovations for detection of obstacles and road using a Multilayer Laserscanner. The classification of pedestrians and other road users require robust object tracking algorithms. Their reliability strongly depends on a precise ego-moti
- PDF / 1,861,998 Bytes
- 25 Pages / 439.414 x 666.172 pts Page_size
- 9 Downloads / 185 Views
Laserscanner Innovations for Detection of Obstacles and Road
U. Lages, IBEO Automobile Sensor GmbH K. Ch. Fuerstenberg, K. C. J. Dietmayer, University of Ulm Abstract This paper presents innovations for detection of obstacles and road using a Multilayer Laserscanner. The classification of pedestrians and other road users require robust object tracking algorithms. Their reliability strongly depends on a precise ego-motion compensation. A new method of ego-motion estimation based on relative velocities of tracked objects will be introduced. This method operates especially under undefined driving conditions, such as skidding of the ego vehicle, where algorithms based on the vehicle’s INS data fail. High performance object tracking and classification algorithms for real-time data processing will be introduced. Especially the results of a new pedestrian recognition technique will be presented. Methods for road detection and road type classification based on Laserscanner data will be explained in a separated chapter. The fusion of road detection and object recognition will enable a better interpretation of driving situations.
1
Introduction
There are more than 1,200,000 accidents in the European traffic with about 1,600,000 injured persons and 42,000 deaths every year [1-4]. Pedestrian accidents - about 160,000 with more than 6,000 killed pedestrians - represents the second largest source of traffic related injuries. Last year object detection and tracking in the nearfield of the ego vehicle using Laserscanners were presented at the AMAA conference [5]. In order to obtain a further improvement of accident prevention and a reduction of both injuries and deaths, innovative safety systems must be integrated into future cars. Laserscanners and their algorithms are able to deliver data relevant for such active safety systems [6,7].
224
Safety
2
Multilayer Laserscanner
The MultiLayer Laserscanner used for environment detection is the latest member of IBEO Automobile Sensor GmbH’s Automotive Laserscanner family (figure 1).
Fig. 1.
Multilayer laserscanner.
Based on the experiences with the previous LD Automotive model, the LD ML uses 4 scan planes instead of one for the observation of the vehicles environment, obtaining a vertical opening angle of 3.2 degree to the front. This allows a compensation for the vehicles pitch angle, as shown in figure 2.
Fig. 2.
Multilayer Laserscanner with four individual scanning planes.
The Laserscanner provide a variable scan frequency from 10 to 40Hz. The angular resolution varies with the scan frequency, at 10Hz the angular resolution is 0.25 degree. The horizontal scan area is up to 270 degree and the object range up to 100m. More Laserscanner details and a description of the working principle can be found in [7,9].
Laserscanner Innovations for Detection of Obstacles and Road
3
Ego-Motion Estimation
The Laserscanner measures relative distances to objects in the scan area. Consequently, the Laserscanner is able to calculate only the relative speed of all objects. However, b
Data Loading...
