Environment Detection

Environment detection

Predictive sensor technology (e.g. radars) enable more adaptive activation of airbag systems shortly before a collision. The prerequisite is a reliable and quick detection of the vehicle surroundings in all traffic situations - even under adverse environmental influences. The challenge is to maximise accuracy while minimizing sensor cycle times, taking into account weather influences on sensors (e.g. rain, fog) and in highly dynamic driving situations (e.g. skidding before a collision). To this end, new methods of signal processing are being researched that determine the crash-relevant parameters of an impending accident with another vehicle or a pedestrian.

Wheel detection

The steadily increasing resolution of environment sensors such as radar sensors enables a continuous increase in object detection quality and details. Especially in the pre-crash phase, when the distance between sensors and potential collision objects is small, additional information can be collected and evaluated. Particularly advantageous are object features whose position is statically located within the object, e.g. the position of the vehicle tyres. These distinctive points enable reliable object detection and tracking.

Therefore, CARISSMA is developing a method to determine the position and speed of rotating vehicle tyres based on the micro-Doppler effect. The additional information is fused with camera and lidar sensor data and provides a robust and extended description of the detected objects.

Contact: Prof. Dr.-Ing. Thomas Brandmeier, Dr. Dagmar Steinhauser

Robust against environmental influences

In order to increase safety and comfort through driver assistance systems up to automated driving, the reliable use of environment sensors is absolutely essential in any environmental condition. In CARISSMA the performance of sensor systems is tested under reproducible boundary conditions in a defined test environment by using weather systems. For the different sensor types (camera, radar, lidar) different disturbing effects are determined and characterised depending on the environmental influences (rain, fog, light). Methods to reduce disturbance effects on sensor systems as well as models for (sensor) disturbance variables are developed to enable an optimal perception of the environment.

Contact: Prof. Dr.-Ing. Thomas Brandmeier, Dr. Dagmar Steinhauser

Robust object tracking

Almost every 6th accident of a passenger car with injured passengers shows a skidding in the pre-crash phase despite modern ESP systems (source: GIDAS). The causes of this are, for example, inappropriate driver interventions in safety-critical driving situations or a slippery road surface caused by snow and ice. Skidding is a special, non-linear vehicle movement which is strongly dependent on external environmental conditions (e.g. friction coefficient of the road surface).

In order to be able to activate future safety systems by predictive sensors in these driving conditions, the close vehicle environment must be reliably and continuously detected. In addition, the activation of passive safety systems (e.g. airbags) represents a critical decision and is classified among the highest Automotive Safety Integration Level (ASIL). To increase robustness and safety, independent plausibility check methods of the environment information from predictive sensors are therefore developed and researched with new methods of object detection and tracking in critical driving situations, implemented on prototype vehicles and tested under real environmental conditions in CARISSMA.

Contact: Prof. Dr.-Ing. Thomas Brandmeier, Robert Naumann

Pedestrian detection

Approximately half of all road deaths worldwide are vulnerable road users (VRUs) such as pedestrians or cyclists. VRUs are always the weaker collision partner in road traffic. The detection of VRUs by means of sensor technology is therefore important to provide the driver with optimal support and thus increase the safety of all road users. In particular with regard to autonomous driving, robust detection is essential in any traffic situation. This must also function reliably in unclear, complex scenarios as well as in the event of disturbances such as bad weather conditions.

In CARISSMA radar-based methods are investigated to enable the robust and fast classification of VRUs. High-resolution radar sensors for the near field are used for this purpose. By evaluating the micro Doppler effect (speed distribution due to movements of the extremities) and high-resolution distance information, important insights into the behaviour of the VRU can be gained in addition to the refined classification. Early indicators are examined in order to register a change in the motion sequence (e.g. change of direction) at an early stage. Based on this, methods are developed to include them in a path prediction. This is particularly important, since the behaviour of the pedestrian is decisive for whether a dangerous situation arises. The path prediction for all relevant objects is therefore important in order to be able to react early in a dangerous situation.

Contact: Prof. Dr.-Ing. Thomas Brandmeier, Patrick Held

Number plate recognition

The THI License Plate Dataset (TLPD) was originally created to support the development of an algorithm capable of anonymising license plates collected with a dashboard camera. Today, the data set contains more than 17,000 vehicle images and approximately 18,000 labelled number plates taken at various angles, distances, lighting and weather conditions. The dataset is accessible free of charge and can only be used for research purposes.

For more information and to download the data set, please follow the link: www.thi.de/go/license-plate-detection

Contact person: Prof. Dr. Alessandro Zimmer