- Automated driving
Using modern augmented reality methods, driver assistance systems can be qualitatively tested with significantly reduced effort. In this way, EVOMOTIV contributes to improving quality and reducing the effort involved in testing camera-based ADAS.
Project Details
Initial situation:
Camera systems are increasingly finding their way into today's automobiles. They recognize lanes, traffic signs, other road users and pedestrians. They increase the safety and comfort of the driver. Due to their complexity, careful testing of such systems is particularly important.
Project assignment:
EVOMOTIV wants to contribute to making driver assistance systems safer. For this reason, a testing system is to be developed that can test camera-based driver assistance systems in a simple, realistic, and reproducible way. This is intended to improve the quality of testing while simultaneously reducing effort.
Implementation:
Based on modern augmented reality methods, the camera data recorded in the vehicle will be expanded with objects relevant to testing. The expanded scenarios will then be forwarded in real time to the evaluating control unit. This should react to the test scene in the same way as to the corresponding real scene.
Solution:
Lane departure warning systems, for example, can be qualitatively tested with such an AR-FAS test system with significantly reduced effort. Curves with any radius can be used reproducibly just as easily as lane widening, narrowing or overlaying. This can increase the quality of the overall system.
Camera systems are increasingly finding their way into today's automobiles. They recognize lanes, traffic signs, other road users and pedestrians. They increase the safety and comfort of the driver. Due to their complexity, careful testing of such systems is particularly important.
Project assignment:
EVOMOTIV wants to contribute to making driver assistance systems safer. For this reason, a testing system is to be developed that can test camera-based driver assistance systems in a simple, realistic, and reproducible way. This is intended to improve the quality of testing while simultaneously reducing effort.
Implementation:
Based on modern augmented reality methods, the camera data recorded in the vehicle will be expanded with objects relevant to testing. The expanded scenarios will then be forwarded in real time to the evaluating control unit. This should react to the test scene in the same way as to the corresponding real scene.
Solution:
Lane departure warning systems, for example, can be qualitatively tested with such an AR-FAS test system with significantly reduced effort. Curves with any radius can be used reproducibly just as easily as lane widening, narrowing or overlaying. This can increase the quality of the overall system.
ProjectTools
Unity
Blender
OpenCV
C#
C++
