Hidden Targets in ADAS testing

220_HiddenTargetEvaluating radar or LIDAR performance for advanced driver assistance systems can be difficult when targets can be hidden behind each other. Using the new polygon targets feature of the RT-Range makes it much easier to analyse sensor data from hidden targets and benchmark performance.

If advanced driver assistance systems are going to help with pedestrian avoidance then it is essential to know how quickly the sensors can correctly identify the pedestrian. As a pedestrian steps out from behind a parked car, the RT-Range can precisely measure when the pedestrian became visible, or even the percentage of the pedestrian that is visible. This can be critical for the development of radar, LIDAR and vision systems where false alarms can be annoying to the driver or even dangerous if the vehicle is to avoid the pedestrian automatically.

Setting up a polygon in the RT-Range is simple. Up to 8 points can be used to represent the vehicle as a 2D, flat object. There is no need to represent the vehicle using thousands of 3D points, which would be difficult to measure. Polygons can be saved to files and then recalled quickly. The position of the RT Inertial and GPS Navigation system in the vehicle can be moved independently from the polygon so that it is not necessary to put it in the vehicle in the same place each time.

The CAN bus output includes the forward and lateral range to the closest point on the polygon; the instantaneous velocity of the polygon point compared to the hunter’s sensor; information on which points on the polygon are interpolated to when the closest point was found and a measurement of the percentage of the polygon that can be seen. Real-time measurements can be displayed on a laptop. Data can be post-processed so that changes to the configuration can be applied retrospectively.

The polygon outputs use the same base technology that all the outputs in the RT-Range have. Wireless LAN is used to transmit all the data from the target vehicles to the hunter vehicle with minimal delay. Full motion of each vehicle (including roll, pitch, heading, acceleration and angular rates) is included in the calculations. By using the RT Inertial and GPS Navigation System the RT-Range is aware of the difference between track angle and heading, so the RT-Range works correctly in corners and when stationary. All the measurements from all the vehicles are available in the Hunter vehicle and can be captured using CAN.

For more information on how the RT-Range can help you develop your ADAS solution please contact OxTS.