In some examples, a radar system is configured to mount on an ownship vehicle for interleaving a weather detection mode and an object detection mode. The radar system comprises a phased-array radar device configured to receive weather signals in the weather detection mode, receive sensing signals in the object detection mode, and interleave the weather detection mode and the object detection mode. The radar system further comprises processing circuitry configured to determine weather conditions based on the received weather signals and detect an object based on the received sensing signals.

Disclosed are techniques for employing deep learning to analyze radar signals. In an aspect, an on-board computer of a host vehicle receives, from a radar sensor of the vehicle, a plurality of radar frames, executes a neural network on a subset of the plurality of radar frames, and detects one or more objects in the subset of the plurality of radar frames based on execution of the neural network on the subset of the plurality of radar frames. Further, techniques for transforming polar coordinates to Cartesian coordinates in a neural network are disclosed. In an aspect, a neural network receives a plurality of radar frames in polar coordinate space, a polar-to-Cartesian transformation layer of the neural network transforms the plurality of radar frames to Cartesian coordinate space, and the neural network outputs the plurality of radar frames in the Cartesian coordinate space.

Disclosed are techniques for employing deep learning to analyze radar signals. In an aspect, an on-board computer of a host vehicle receives, from a radar sensor of the vehicle, a plurality of radar frames, executes a neural network on a subset of the plurality of radar frames, and detects one or more objects in the subset of the plurality of radar frames based on execution of the neural network on the subset of the plurality of radar frames. Further, techniques for transforming polar coordinates to Cartesian coordinates in a neural network are disclosed. In an aspect, a neural network receives a plurality of radar frames in polar coordinate space, a polar-to-Cartesian transformation layer of the neural network transforms the plurality of radar frames to Cartesian coordinate space, and the neural network outputs the plurality of radar frames in the Cartesian coordinate space.

A system for navigating a host vehicle may receive an image representative of an environment of the host vehicle and determine a planned navigational action for accomplishing a navigational goal of the host vehicle. The system may identify a target vehicle, determine a current speed of the target vehicle, and assume a maximum braking rate capability of the target vehicle. The system may determine a next-state distance between the host vehicle and the target vehicle that would result if the planned navigational action was taken. The system may implement the planned navigational action if the host vehicle may be stopped using a predetermined sub-maximal braking rate within a distance that is less than the determined next-state distance summed together with a target vehicle travel distance determined based on the current speed of the target vehicle and the maximum braking rate capability of the target vehicle.

A method in a time switched multiple input and multiple output (MIMO) radar system comprising, receiving (610) from an antenna array a plurality of data points representing a radar signal reflected from plurality of objects, forming (620) a first set of beams from the plurality of data points, wherein the first set of beams are making a first set angles with a normal to the antenna array, detecting a set of objects (410A-L) from the first set of beams, determining (630) a set of Doppler frequencies of the set of objects, computing (650) a self-velocity representing a velocity of the antenna array from the set of Doppler frequencies and the first set of angles, and correcting (660) the plurality of data points using the self-velocity and a second set of angles to generate plurality of corrected data points.

A moving object recognition apparatus includes an object detection section, a position detection section, a road direction estimation section, and a moving direction estimation section. The object detection section detects a moving object that moves on a road around an own vehicle and a roadside object by the road, from objects present around the own vehicle. The position detection section detects positions of the moving object and the roadside object detected by the object detection section. The road direction estimation section estimates a road direction of the road on which the moving object is moving, based on the position of the roadside object detected by the position detection section. The moving direction estimation section estimates a moving direction of the moving object based on the road direction estimated by the road direction estimation section.

A moving object recognition apparatus includes an object detection section, a position detection section, a road direction estimation section, and a moving direction estimation section. The object detection section detects a moving object that moves on a road around an own vehicle and a roadside object by the road, from objects present around the own vehicle. The position detection section detects positions of the moving object and the roadside object detected by the object detection section. The road direction estimation section estimates a road direction of the road on which the moving object is moving, based on the position of the roadside object detected by the position detection section. The moving direction estimation section estimates a moving direction of the moving object based on the road direction estimated by the road direction estimation section.

A system for a host vehicle includes a processor programmed to receive, from an image capture device, an image representative of an environment of the host vehicle, detect at least one obstacle in the environment of the host vehicle based on an analysis of the at least one image, determine a velocity of the host vehicle and a predicted path for the host vehicle, monitor a driver input to at least one of a throttle control, a brake control, or a steering control associated with the host vehicle, and determine whether the driver input would result in the host vehicle navigating within a proximity buffer relative to the at least one obstacle, wherein the proximity buffer is determined based on the determined velocity, a maximum acceleration capacity of the host vehicle, and a maximum braking capacity of the host vehicle, and a reaction time associated with the host vehicle.

A navigational system for a host vehicle may comprise at least one processor. The processor may be programmed to receive an image representative of an environment of the host vehicle; analyze the image to identify a navigational state associated with the host vehicle; and determine, based on the navigational state, a navigational action for the host vehicle based on a policy that maps possible navigational actions to sensed states. The navigational action may be based on a safety constraint applicable to the navigational state, the safety constraint including a safety distance constraint associated with the host vehicle, wherein the safety distance constraint is based on a determined speed of the host vehicle and a determined speed of a detected target object. The processor may cause an adjustment of a navigational actuator of the host vehicle to implement the determined navigational action.

In an azimuth estimation device, a center generation unit configured to generate, for each peak bin extracted by the extraction unit, a center matrix which is a correlation matrix obtained using values of the same peak bin collected from all of transmitting/receiving channels. A surrounding generation unit is configured to generate, for each of one or more surrounding bins of each of the peak bins, a surrounding matrix which is a correlation matrix obtained using values of the same surrounding bin collected from all of the transmitting/receiving channels. An integration unit is configured to generate, for each peak bin, an integrated matrix which is a correlation matrix obtained by weighting and adding the center matrix and the one or more surrounding matrices. An estimation unit is configured to execute an azimuth estimation calculation using the integrated matrix generated by the integration unit.