A method of protecting humans in an environment of a moving machine is provided that comprises the environment being monitored by means of a protective device that is configured to detect one or more kinematic parameters of a respective object located in the environment and controlling the moving machine in dependence on detected kinematic parameters of the respective object to initiate a protective measure. The protective equipment here detects the polarization properties and a movement modulation of the respective object in dependence on which the respective object is classified with respect to whether the respective object is a human. In particular only when the respective object was classified as a human, the protective equipment controls the moving machine to initiate the protective measure in dependence on detected kinematic parameters of this respective object.

The present disclosure provides a system for processing radar data. The system may comprise a frequency generator configured to generate a reference frequency signal; a timing module configured to generate a shared clock signal or a plurality of timing signals; and a plurality of radar modules in communication with the frequency generator and timing module. The radar modules may be configured to: (i) receive the reference frequency signal and at least one of a shared clock signal and a timing signal, (ii) transmit a first set of radar signals based in part on the reference frequency signal and at least one of the shared clock signal and the timing signal, and (iii) receive a second set of radar signals reflected from a surrounding environment. The system may comprise a processor configured to process radar signals received by each radar module, by coherently combining radar signals using phase and timestamp information.

This document describes techniques, apparatuses, and systems for sensor fusion for object-avoidance detection, including stationary-object height estimation. A sensor fusion system may include a two-stage pipeline. In the first stage, time-series radar data passes through a detection model to produce radar range detections. In the second stage, based on the radar range detections and camera detections, an estimation model detects an over-drivable condition associated with stationary objects in a travel path of a vehicle. By projecting radar range detections onto pixels of an image, a histogram tracker can be used to discern pixel-based dimensions of stationary objects and track them across frames. With depth information, a highly accurate pixel-based width and height estimation can be made, which after applying over-drivability thresholds to these estimations, a vehicle can quickly and safely make over-drivability decisions about objects in a road.

A driving support device includes a storage device storing a program and a hardware processor. The hardware processor executes the program stored in the storage device to perform driving support of a vehicle based on a detection result of at least one of a radar device and an imaging device mounted in the vehicle, determine whether the vehicle is traveling in an underpass which is a traffic route along which the vehicle is able to pass under an overlying structure, and suppress an operation of the driving support when the vehicle is determined to be traveling under the underpass.

In a distance detection apparatus for a vehicle, a distance detection unit is configured to detect object distances to an object around the vehicle by transmitting and receiving radar waves. A detection distance estimation unit is configured to estimate an estimated detection distance having a maximum value of the object distance, based on the object distances and intensities of the radar waves reflected by the object. A lost-track distance estimation unit is configured to estimate, as a lost-track distance, the object distance at a timing of transition from a detected state where the object is recognized to a non-detected state where the object is not recognized, based on a result of detection by the distance detection unit. A performance determination unit is configured to determine a degradation of detection performance of the distance detection unit based on the estimated detection distance and the lost-track distance.

This document describes techniques and systems to detect and localize NLOS objects using multipath radar reflections and map data. In some examples, a processor of radar system can identify a detection of an object using reflected EM energy and determine, using map data, whether a direct-path reflection associated with the detection is within a roadway. In response to determining that the direct-path reflection is not located within the roadway, the processor can determine whether a multipath reflection (e.g., a multipath range and multipath angle) associated with the detection is viable. In response to determining that the multipath reflection is viable, the processor can determine that the detection corresponds to an NLOS object. The processor can also provide the NLOS object as an input to an autonomous or semi-autonomous driving system of the vehicle, thereby improving the safety of such systems.

Motion-related 3D feature extraction by receiving a set of sequential radar signal data frames associated with a subject, determining radar signal amplitude and phase for each data frame, determining radar signal phase changes between sequential data frames, and extracting, by a trained machine learning model, one or more three-dimensional features from the sequential radar signal data frames according to the radar signal amplitude and the radar signal phase changes between sequential data frames.

A method for detecting a moving target through a vehicle radar system includes acquiring radar data from a radar sensor disposed with respect to a vehicle, and using the radar data to detect a plane of a structure. The method also includes setting a reference classification line based on the detected plane, and determining whether the moving target is in a line of sight (LOS) area or a non-line of sight (NLOS) area based on the reference classification line.

Method for detecting traffic congestion using a motor vehicle radar system, comprising multi-beam radar sensors (21-24) in the rear and front corners of the vehicle, the method comprising the steps of: dividing the radar sensors (Df_l, Df_r, Dr_l, Dr_r) into four angular sectors (Zfront, Zrear, Zleft, Zright) extending to the front, to the rear, to the right and to the left of the vehicle respectively, selecting for each angular sector, from the beams for which no target is detected, the (Dfront, Drear, Dleft, Dright) beam having the shortest reach distance, detecting the amplitude of reflected beams corresponding respectively to the selected beams, and—analysing the period during which the amplitude of the reflected beams is maintained relative to a predefined time threshold for each angular sector respectively, the method detecting a traffic congestion situation when the analysing step determines simultaneously for the four angular sectors that the period during which the amplitude of the reflected beams is maintained is greater than or equal to the predefined time threshold.

Disclosed are devices, systems and methods for compensating radar measurements of a vehicle. One exemplary method includes generating a set of velocity hypotheses of a target object based on a first measurement data obtained from sensors mounted on the vehicle; generating cluster velocity estimates by applying a clustering algorithm to a second measurement data obtained from the sensors; and providing one or more selected velocity hypotheses from the set of velocity hypothesis as compensated radar measurements for the target object based on the cluster velocity estimates.