Wheel extraction based on micro doppler distribution using high-resolution radar

Abstract

With the advent of advanced driver assistant systems (ADAS) in urban scenarios, a fast and reliable classification and motion estimation of wheel-based vehicles such as cars, trucks or motorcycles is crucial. The fact that the wheels' velocities differ from the vehicle's chassis velocity is exploited. For the first time, a fully automated approach based on the Doppler distribution extracts the exact positions of the wheels. The Normalized Doppler Moment is calculated, describing the Doppler signature of each reflection based on the Doppler distributions of wheels. Locations with high values reveal the positions of the wheels. Besides the classification, the vehicle's orientation and therefore the driving direction can be estimated. Furthermore the position of the rear axle is estimated, which is essential for a reliable prediction of rotational movements and yaw rate estimation. Experimental results with a 77 GHz automotive radar sensor demonstrate the feasibility of the approach.

Doppler-Spectrum Feature-Based Human–Vehicle Classification Scheme Using Machine Learning for an FMCW Radar Sensor

“…While it can be observed that the Doppler spectra of a moving vehicle are similar, given the sharp shape shown in Figure 8a, those of the walking human fluctuate due to the non-rigid motion, as shown in Figure 8b. However, the micro-Doppler component from the wheel is quite small in most cases compared to the overall gross motion of the Doppler form [27]. Moreover, if using a radar sensor with multiple receive channels, we can use the angular spectrum based on DBF (digital beam forming).…”

“…However, the micro-Doppler component from the wheel is quite small in most cases compared to the overall gross motion of the Doppler form [27]. Moreover, if using a radar sensor with multiple receive channels, we can use the angular spectrum based on DBF (digital beam forming).…”

Recent evolution of automotive imaging radar and its information content

“…It can be clearly distinguished between the bulk of the vehicle, where the radial velocity follows a smooth transition from top left to bottom right, and the wheels, where the radial velocity evidently deviates from this pattern. Kellner et al [8] proposed a method how such wheel positions can be extracted from the radar image. The resulting information of the wheel positions can be used to conclude onto the target vehicles orientation and thereby support tracking.…”

Instantaneous Actual Motion Estimation with a Single High-Resolution Radar Sensor

“…To correctly set the reference point, the moving direction of the vehicle and the offset of the reference point to the backside of the vehicle have to be known. The offset has be be assumed based upon classification or by trying to identify the wheel positions [7], or -in this case -via a priori knowledge. For nonlinear motion, the information about the moving direction cannot be directly gained, without first knowing the turning direction and the segment of the turn.…”