A radar signal target detection and display system employs feedback to produce improved detection and tracking. Portions of the radar images for each antenna channel to search for the target of interest are defined using the track state to define the center of the search region and using the track covariance matrix to define the size of the region. Moving reference processing (MRP) is performed. MRP employs a search centered on the motion state derived from the detection sent to the tracker on the previous coherent processing interval (CPI). Space-time adaptive processing (STAP) is employed on each CPI using a unique set of adaptive degrees of freedom (DOFs) derived from pre-MRP and post-MRP complex radar image amplitudes for each antenna channel.

Error that occurs when an absolute velocity of a target object is measured by using an antenna installed on a ship body that rocks and drifts in a complex manner since it floats on the sea is reduced. An antenna is installed on a ship body and transceives electromagnetic waves. A roll angle and a pitch angle of the ship body are detected by using an inclination sensor. An antenna velocity calculator calculates an antenna velocity of the antenna by using the detected roll and pitch angles of the ship body. An antenna velocity compensator compensates the antenna velocity of the antenna for a relative velocity between the ship body and a target object, the antenna velocity calculated by the antenna velocity calculator, the relative velocity obtained based on reflection waves received by the antenna.

A system is disclosed for Doppler nulling configured for security. The system may include a receiver or transmitter node. The receiver or transmitter node may include a communications interface with an antenna element and a controller. The controller may include one or more processors and have information of own node velocity and own node orientation relative to a common reference frame. The receiver or transmitter node may be time synchronized to apply Doppler corrections to signals, the Doppler corrections associated with the receiver or transmitter node's own motions relative to the common reference frame, the Doppler corrections applied using Doppler null steering along Null directions based on a protocol. The protocol may include a protocol modulation, such as a modulation of the signals for security purposes.

A roll angle determination method for an object in flight, the method comprising generating and transmitting signal patterns and analyzing a detected signal at a sensor provided on the object and determining a roll angle and zero time of a fundamental frequency of the transmitted signal on a transmitter clock at a sensor processor clock time, and the sensor receiver synchronizing its time with the polarized RF scanning reference source time.

A system and a method of generating a three-dimensional terrain model using one-dimensional interferometry of a rotating radar unit is provided herein. Height information is evaluated from phase differences between two echoes by applying a Kalman filter in relation to a phase confidence map that is generated from phase forward projections relating to formerly analyzed phase data. The radar system starts from a flat earth model and gathers height information of the actual terrain as the platform approaches it. Height ambiguities are corrected by removing redundant 2 multiples from the unwrapped phase difference between the echoes.

A system for signal processing is provided that obviates the use of prior-knowledge, such as synthetic aperture radar (SAR) imagery, in time compressed signal processing (i.e. it can be knowledge unaided). The knowledge-unaided power centroid (PC.sub.KU) is found by evaluating a covariance matrix R.sub.SCM for its moments m.sub.i. Because R.sub.SCM uses a sample signal, rather than SAR data, the power centroid PC.sub.KU may be found without needing SAR data.

A radar system comprises a plurality of receive antennas that receive a radar signal. One or more processors are configured to generate an n-dimensional point cloud comprising values for parameters of an object in an environment of the radar system, where the n-dimensional point cloud is generated based upon the radar signal, and where n is greater than 2. The one or more processors are further configured to generate from the n-dimensional point cloud a 2D RGB image representing the values of the parameters in different colors, respectively. The parameters can comprise at least object radar cross-section, height, and velocity, etc. The one or more processers are further configured to provide the 2D RGB image to a convolutional neural network that assigns a classification to the object based on the 2D RGB image.

Provided is a method for movement estimation and movement compensation of a target object that can be applied without introducing restrictions on antenna placement. The present invention provides a radar apparatus including: a radar signal transmission-reception unit acquiring a radar signal acquired by measurement using a transmission antenna and a reception antenna, and a measurement time of the radar signal; a velocity candidate control unit holding a setting of a velocity candidate set of a target object; a velocity estimation imaging unit generating a radar image applied with movement compensation by using each velocity candidate; a velocity estimation unit selecting an estimated velocity from a velocity candidate set, based on comparison of each generated radar image; and an output image imaging unit generating a final output image applied with movement compensation using an estimated velocity.

In various examples, a deep neural network(s) (e.g., a convolutional neural network) may be trained to detect moving and stationary obstacles from RADAR data of a three dimensional (3D) space, in both highway and urban scenarios. RADAR detections may be accumulated, ego-motion-compensated, orthographically projected, and fed into a neural network(s). The neural network(s) may include a common trunk with a feature extractor and several heads that predict different outputs such as a class confidence head that predicts a confidence map and an instance regression head that predicts object instance data for detected objects. The outputs may be decoded, filtered, and/or clustered to form bounding shapes identifying the location, size, and/or orientation of detected object instances. The detected object instances may be provided to an autonomous vehicle drive stack to enable safe planning and control of the autonomous vehicle.

In various examples, a deep neural network(s) (e.g., a convolutional neural network) may be trained to detect moving and stationary obstacles from RADAR data of a three-dimensional (3D) space, in both highway and urban scenarios. RADAR detections may be accumulated, ego-motion-compensated, orthographically projected, and fed into a neural network(s). The neural network(s) may include a common trunk with a feature extractor and several heads that predict different outputs such as a class confidence head that predicts a confidence map and an instance regression head that predicts object instance data for detected objects. The outputs may be decoded, filtered, and/or clustered to form bounding shapes identifying the location, size, and/or orientation of detected object instances. The detected object instances may be provided to an autonomous vehicle drive stack to enable safe planning and control of the autonomous vehicle.