Lateral velocity estimation based on automotive radar sensors

Abstract

Automotive radar sensors are applied to measure the target range, azimuth angle and radial velocity simultaneously even in multiple target situations. The single target measured data are necessary for target tracking in advanced driver assistance systems (ADAS) e.g. in highway scenarios. In typical city traffic situations the radar measurement is also important but additionally even the lateral velocity component of each detected target such as a vehicle is of large interest in this case. It is shown in this paper that the lateral velocity of an extended target can be measured even in a mono observation situation. For an automotive radar sensor a high spectral resolution is required in this case which means the time on target should be sufficiently large.

Extending Reliability of mmWave Radar Tracking and Detection via Fusion With Camera

“…Some papers [14], [15], [20], [21] illustrated the alignment between radars and other heterogeneous sensors. They assumed radars and other sensors are working at identical principle.…”

“…However, the errors are not corrected during detections. In [21], Folster and Rohling introduced a lateral velocity estimation for automotive radar sensors for ADAS. The method simply refers to the location of the target on the actual detections.…”

“…It represents the relative weight of the measurements compared to the current state estimate, and is typically changing over time as the noise keeps changing throughout time. Rewriting the Equation (21) with Equation (20):…”