Stereoscopic radar

Abstract

A computer-implemented method for detecting one or more objects using at least two radar sensors (310; 312) of a stereoscopic radar assembly (314) arranged spaced apart from each other. The method includes receiving first radar signal data and second radar signal data determined using a first radar sensor (310) and a second radar sensor (312), respectively. Both radar signal data describe characteristics of a respective radar signal acquired using one of the two radar sensors (310; 312). The first radar signal data and the second radar signal data are used to determine one or more combinations of distance and radial velocity including intensity peaks. For one or more of the respective combinations of distance and radial velocity, first and second frequency spectrums (302; 304) describing intensities as a function of azimuth angle and elevation angle are determined using the first and second radar signal data, respectively. One or more positions of the one or more objects in azimuth and elevation angles are determined by matching and comparing the first and second spectra (302; 304).

Description

Technical Field

[0001] This invention relates to the field of radar detection. More specifically, this invention relates to a computer-implemented method for detecting one or more objects using at least two radar sensors arranged spaced apart from each other in a stereo radar assembly. Background Technology

[0002] Radar sensing is a holistic solution, for example, used in autonomous driving, as it can be used to generate the position and radial velocity of objects around a vehicle (e.g., a car). The ability of radar to distinguish two closely spaced objects depends on, for example, the number of antennas, which can be limited by cost and size constraints. Multiple-input multiple-output (MIMO) radar employs a virtual array concept to increase the radar aperture without increasing the number of physical antennas. However, MIMO radars compatible with automotive constraints still suffer from, for example, low spatial resolution. In addition to low spatial resolution, a single radar may be unable to detect objects with low signal-to-noise ratio (SNR) as well as occluded targets. Therefore, a method for improved object detection is needed.

[0003] The object of this invention is to provide a computer-implemented method, computer program, and computer apparatus for detecting one or more objects using at least two radar sensors arranged spaced apart from each other in a stereo radar assembly. This object is achieved by the features of the independent claims. Summary of the Invention

[0004] In one aspect, a computer-implemented method is disclosed for detecting one or more objects using at least two radar sensors arranged spaced apart from each other in a stereo radar assembly. The method includes receiving first radar signal data determined using a first radar sensor of the at least two radar sensors. The first radar signal data describes characteristics of the first radar signal acquired using the first radar sensor. Second radar signal data determined using a second radar sensor of the at least two radar sensors is received. The second radar signal data describes characteristics of the second radar signal acquired using the second radar sensor. Using the first radar signal data, one or more first combinations of range and radial velocity are determined, for which the first radar signal data includes an intensity peak. Using the second radar signal data, one or more second combinations of range and radial velocity are determined, for which the second radar signal data includes an intensity peak. For one or more of the determined one or more first and second combinations of range and radial velocity, the first radar signal data is used to determine a first spectrum describing the intensity as a function of azimuth and elevation. Furthermore, the second radar signal data is used to determine a second spectrum describing the intensity as a function of azimuth and elevation. Furthermore, the first and second spectra are used to determine one or more positions of one or more objects in terms of azimuth and elevation. The determination includes matching and comparing the first spectrum and the second spectrum.

[0005] This method enables the efficient implementation of stereo radar, thereby improving radar system performance. It does not require extensive adaptation of the radar system's software or hardware. The at least two radar sensors discussed can be two radar sensors arranged spaced apart from each other. In the following text, such a combination of at least two radar sensors whose detection ranges at least partially overlap is referred to as a stereo radar assembly. For example, the at least two radar sensors can be independent of each other, i.e., asynchronous or only coarsely synchronized.

[0006] Each of the at least two radar sensors can operate, for example, in monostatic and / or bistatic modes. The at least two radar sensors can also operate, for example, in multistatic mode. In monostatic mode, the radar sensor operates as a receiver for receiving reflected radar signals emitted by the same radar sensor; that is, the receiver and transmitter are co-located as respective radar sensors. In bistatic mode, the radar sensor operates as a receiver for receiving reflected radar signals emitted by other radar sensors among the at least two radar sensors; that is, the receiver and transmitter are arranged spaced apart from each other. In multistatic mode, in the case of a combination of more than two radar sensors, the radar sensor operates as a receiver for receiving reflected radar signals emitted by multiple other radar sensors. In the case of a component comprising n>2 sensors, the sensor of the component operating in multistatic mode can, for example, be used as a receiver for receiving radar signals emitted by up to n-1 other radar sensors of the component.

[0007] Each radar sensor in a stereo radar assembly can, for example, transmit radar signals. These radar signals can interact with an object, such as being reflected, scattered, and / or diffracted, and are received by the transmitting radar sensor and / or one or more other radar sensors in the assembly due to these interactions.

[0008] In monostatic mode, a single radar sensor acts as both a transmitter and receiver of radar signals. The radar sensor transmits radar signals and receives reflections of the transmitted signals. Using the reflected radar signals acquired by the radar sensor, the distance, velocity, and / or other characteristics of the reflecting object can be determined. When operating only in monostatic mode, first radar signal data can describe the strength of a first radar signal received by a first radar sensor, which was transmitted by the first radar sensor. When operating only in monostatic mode, second radar signal data can describe the strength of a second radar signal received by a second radar sensor, which was transmitted by the second radar sensor.

[0009] In bistatic mode, different radar sensors arranged spaced apart from each other act as transmitters and receivers of radar signals. The transmitter radar sensor transmits a radar signal that interacts with an object, such as being reflected, scattered, and / or diffracted. The receiver radar sensor, arranged spaced apart from the transmitter radar sensor, receives the radar signal generated due to the interaction. When operating only in bistatic mode, first radar signal data can describe the strength of a first radar signal received by a first radar sensor, wherein the received first radar signal was transmitted by a second radar sensor. When operating only in bistatic mode, second radar signal data can describe the strength of a second radar signal received by a second radar sensor, wherein the received second radar signal was transmitted by the first radar sensor.

[0010] When operating in monostatic and bistatic modes, the first radar signal data can describe a combination of the strengths of first radar signals received by the first radar sensor, wherein the received first radar signals include radar signals emitted by the first radar sensor and radar signals emitted by the second radar sensor. Therefore, the first radar signal data can include a combination of radar signal data describing the received radar signals emitted by the first radar sensor and radar signal data describing the received radar signals emitted by the second radar sensor. Similarly, when operating in monostatic and bistatic modes, the second radar signal data can describe a combination of the strengths of second radar signals received by the second radar sensor, wherein the received second radar signals include radar signals emitted by the first radar sensor and radar signals emitted by the second radar sensor. Therefore, the second radar signal data can include a combination of radar signal data describing the received radar signals emitted by the first radar sensor and radar signal data describing the received radar signals emitted by the second radar sensor.

[0011] In the case of more than two radar sensors (e.g., n radar sensors), the first radar signal data may include combined data describing the received radar signals emitted by n or fewer of the n radar sensors. In the case of more than two radar sensors (e.g., n radar sensors), the second radar signal data may include combined data describing the received radar signals emitted by n or fewer of the n radar sensors.

[0012] For an assembly comprising n>2 radar sensors arranged spaced apart from each other, the method may, for example, include receiving first to nth radar signal data determined using first to nth radar sensors among the n radar sensors. The first to nth radar signal data describes the characteristics of the first to nth radar signals acquired using the first to nth radar sensors. Using the first to nth radar signal data, one or more first to nth combinations of range and radial velocity are determined, for which the first to nth radar signal data includes intensity peaks. For one or more of the determined one or more first to nth combinations of range and radial velocity, the first to nth radar signal data are used to determine first to nth spectra describing the intensity as a function of azimuth and elevation. Furthermore, the first to nth spectra are used to determine one or more positions of one or more objects in terms of azimuth and elevation. The determination includes matching and comparing the first to nth spectra.

[0013] For example, radar signals emitted by at least two radar sensors can have different waveforms in order to, for example, reduce the possibility of interference and improve the coexistence of different radar signals. Different frequency bands can be assigned to different radar sensors. By operating on different frequency bands (i.e., non-overlapping frequency ranges), the possibility of interference between radar signals emitted by different radar sensors can be reduced.

[0014] A stereo radar assembly can be, for example, a stereo radar assembly for a vehicle (especially an automobile). The radar sensors of the stereo radar assembly can be installed, for example, in the vehicle (especially an automobile) and can be configured to detect the environment surrounding the vehicle, i.e., objects in the environment surrounding the vehicle. For example, a stereo radar assembly can include radar sensors for the front, corner, side, and / or rear radars of a vehicle (e.g., an automobile).

[0015] This stereo radar assembly can be used in vehicles, for example, to enable the functions of assisted, automated, and / or autonomous driving systems. For instance, it can be used in automobiles for adaptive cruise control (ACC). Adaptive cruise control is an advanced driver assistance system for road vehicles, configured to automatically adjust the vehicle's speed, for example, to maintain a safe distance from the vehicle ahead. For example, adaptive cruise control can be part of radar-based emergency braking assistance. Furthermore, this stereo radar assembly can be used, for example, for cross-traffic alert (CTA) to detect, for example, cross-traffic behind the vehicle. This cross-traffic alert can work in conjunction with, for example, a blind spot monitoring system and is configured to warn the driver of approaching cross-traffic when reversing out of a parking space.

[0016] Similar to stereo camera systems in the field of optical imaging, stereo radar components can improve radar system performance by combining radar signal data from at least two radar sensors arranged spaced apart from each other. Radar frequencies typically used in vehicles can be, for example, in the range of 76 to 77 GHz, corresponding to a wavelength of approximately 4 mm.

[0017] A radar sensor is a device for emitting radar signals and detecting the reflection of the emitted radar signals from objects within the detection range of the radar sensor. The characteristics of these reflections, and thus the characteristics of the detected radar signals, can depend on the features of the reflecting objects within the detection range of the radar sensor. These features of the objects can include, for example, position, size, shape, surface condition, motion characteristics, and / or trajectory. The reflected radar signals acquired by the radar sensor can, for example, include radar signals emitted by one or more radar sensors. For example, a first radar signal acquired using a first radar sensor can be generated by the reflection of radar signals emitted by a first radar sensor and / or a second radar sensor. A second radar signal acquired using a second radar sensor can, for example, be generated by the reflection of radar signals emitted by a second radar sensor and / or a first radar sensor. The reflected radar signals acquired by the radar sensor can, for example, describe the position (e.g., defined by range, azimuth, and elevation) and radial velocity relative to the radar sensor of one or more detected objects. Therefore, radar signal data describing such radar signals can include four-dimensional information about objects within the detection range of the radar sensor. This four-dimensional information can include the range, azimuth, elevation, and radial velocity of the detected object relative to the detecting radar sensor. The presence of an object within the detection range of a radar sensor can be indicated, for example, by the intensity peaks of the acquired radar signal, and therefore by the intensity peaks of the radar signal data describing the corresponding radar signal. By determining the dependence of the intensity peaks included in the radar signal data on range, azimuth, elevation, and / or radial velocity, the range, azimuth, elevation, and / or radial velocity of the object within the detection range of the radar sensor can be determined.

[0018] Radar signal data may be provided, for example, in the form of acquired radar signals, or radar signal data may be provided, for example, in the form of processed radar signals acquired using radar sensors.

[0019] For example, radar signal data describing radar signals (i.e., detected radar reflections) can be used to determine the range Doppler distribution. The range Doppler profile can be provided as a graphical two-dimensional representation of the intensity of radar reflections received from objects within the detection range of a radar sensor, said radar reflection intensity as a function of range (i.e., distance) and Doppler frequency shift (i.e., the radial velocity of the corresponding object relative to the radar sensor).

[0020] For example, a frequency-modulated continuous wave (FMCW) radar sensor can be used. In the case of an FMCW radar sensor, the transmitted radar signal is frequency-modulated. This frequency modulation allows for range measurement using indirect time-of-flight measurements by comparing the frequency of the received radar signal with a reference (e.g., the transmitted radar signal). Furthermore, the radial velocity can be a measured Doppler shift of the received radar signal. Depending on the relative distance and radial velocity of the object to the radar sensor, the acquired reflected radar signal can include frequency variations. These frequency variations can be processed using suitable techniques such as Fast Fourier Transform (FFT) to extract characteristics of the object, such as range and / or radial velocity. Therefore, by performing an FFT on the acquired reflected radar signal or radar signal data describing the radar signal, the position of the detected object, defined in terms of range, azimuth and / or elevation and / or radial velocity, can be determined. This analysis can, for example, contribute to other applications in target detection, target recognition, target tracking, and autonomous driving systems.

[0021] The range Doppler distributions of an object determined by two radar sensors spaced apart from each other can be different. The range Doppler distribution can describe the intensity spectrum or power spectrum of the radar signal acquired by the radar sensors. In monostatic mode, i.e., when radar sensor data is determined using a radar sensor operating in monostatic mode, the range Doppler distribution can describe the intensity spectrum or power spectrum of the radar signal transmitted and received by the same radar sensor. In bistatic mode, i.e., when radar sensor data is determined using a radar sensor operating in bistatic mode, the range Doppler distribution can describe the intensity spectrum or power spectrum of the radar signal transmitted and received by different radar sensors arranged spaced apart from each other. In a combined monostatic and bistatic mode, i.e., when radar sensor data is determined using radar sensors operating in a combined monostatic and bistatic mode, the range Doppler profile can describe the intensity spectrum or power spectrum of the combined radar signal transmitted and received by the same radar sensor and the radar signal transmitted and received by different radar sensors arranged spaced apart from each other.

[0022] An intensity spectrum or power spectrum describes the distribution of intensity or power over range and radial velocity. The intensity or power spectrum determined by a first radar sensor using at least two radar systems is called the first spectrum, while the intensity or power spectrum determined by a second radar sensor using at least two radar systems is called the second spectrum. Depending on the mode used to determine the first radar signal data, the first spectrum can be, for example, the spectrum of radar signals received and transmitted by the first radar sensor and / or the spectrum of radar signals received by the first radar sensor but transmitted by another radar sensor (e.g., the second radar sensor) arranged spaced apart from the first radar sensor. Similarly, depending on the mode used to determine the second radar signal data, the second spectrum can be, for example, the spectrum of radar signals received and transmitted by the second radar sensor and / or the spectrum of radar signals received by the second radar sensor but transmitted by another radar sensor (e.g., the first radar sensor) arranged spaced apart from the second radar sensor.

[0023] Intensity I and power P are proportional to each other, with area A as a proportionality constant, i.e., I = P / A. Therefore, the intensity spectrum and the power spectrum are proportional to each other.

[0024] To determine the range Doppler distribution, the parameter space spanned by range and radial velocity can be discretized, for example, by dividing it into intervals or bins. Bins can be specified as continuous, non-overlapping intervals of the variables range and radial velocity. Bins can be adjacent and of equal size. Each bin can be assigned an accumulation of radar signal data, which is assigned to the range and velocity included in the corresponding bin. Based on intensity peaks, bins can be identified using radar signal data describing the presence or detection of one or more objects. For these bins, for example, the intensity or power spectrum can be determined as a function of azimuth and elevation angles, corresponding to a two-dimensional image from which the positions of one or more corresponding objects within the area spanned by the azimuth and elevation angles can be determined. This intensity or power spectrum can describe a two-dimensional distribution of radar signal intensity or power, which depends on the azimuth and elevation angles of the radar signal data assigned to the range and velocity included in the corresponding bin. Therefore, the spatial arrangement of objects in terms of azimuth and elevation at distances within the same range interval and at radial velocities within the same radial velocity interval can be determined.

[0025] For example, the presence of one or more objects can be determined by identifying intensity peaks in the first and second spectra at a matching location in terms of azimuth and elevation. By using radar signal data from two radar sensors, higher SNR and / or accuracy can be achieved, and therefore false alarms and / or missed alarms can be avoided. For example, if an object is detectable at a matching location in terms of azimuth and elevation within the radar signal data from both radar sensors, the object's presence can be confirmed. This may even be the case for relatively low intensity peaks that enable the detection of low SNR objects and / or obscured objects. On the other hand, in the absence of a match, intensity peaks can be discarded to avoid false alarms.

[0026] When matching the first and second spectra, the azimuth and elevation angles can be coordinates in a global coordinate system used to locate the position of an object relative to the stereo radar assembly and thus, for example, relative to a vehicle. For each individual radar sensor, the azimuth and elevation angles can be defined in a local coordinate system relative to the corresponding radar sensor. These local coordinate systems can be different from each other because they are arranged to be spaced apart from each other. By transforming these local coordinates into a global coordinate system, the position can be jointly defined relative to a stereo radar assembly comprising two radar sensors.

[0027] For example, the first and second combinations of distance and radial velocity comprise combinations of distance and radial velocity intervals distributed according to a predefined interval distribution. These intervals are specified as bins, which can be specified as continuous, non-overlapping ranges of values. The range of parameter values ​​(i.e., distance and radial velocity) can be divided into a series of intervals. Different combinations of distance and radial velocity can fall into different intervals. Therefore, the range of parameter values ​​can be discretized.

[0028] The parameter space for distance and radial velocity can be discretized, for example, by dividing it into intervals or bins. Based on intensity peaks, bins in which one or more objects are present or detected can be identified. For these bins, for example, the intensity or power spectrum can be determined as a function of azimuth and elevation angles. This intensity or power spectrum of a particular bin corresponds to a two-dimensional image from which the location of one or more corresponding objects within a region spanned by the azimuth and elevation angles can be determined.

[0029] For example, the first interval distribution used to determine a first combination of range and radial velocity is offset relative to the second interval distribution used to determine a second combination of range and radial velocity. Therefore, the bins defined for analyzing the first radar signal data can, for example, be shifted relative to the bins defined for analyzing the second radar signal data. Consequently, due to this shift, the boundary between the two bins defined for the first radar sensor can correspond to a combination of parameter values ​​within the bin defined for the second radar sensor. For example, if the bin is shifted by half its width, it can correspond to the middle of the bin defined for the second radar sensor. This shift can result in slightly different allocations of radar signal data to bins used for different radar sensors. For example, in the case of the second radar sensor, radar signal data allocated to two different adjacent bins defined for the first radar sensor can be allocated to the same bin. This can further improve object detection because the effects arising solely from the choice of bin boundaries can be compensated for and thus avoided.

[0030] For example, a global coordinate system describing azimuth and elevation angles is used to determine the first and second spectra. The use of the global coordinate system involves converting first local coordinates assigned to a first local coordinate system for the first radar sensor and second local coordinates assigned to a second local coordinate system for the second radar sensor into global coordinates in the global coordinate system.

[0031] Transforming local coordinates to a global coordinate system allows for the definition of a common volume sensed by two radar sensors. This volume can be, for example, a 3D space spanned using common distance, azimuth, and elevation as coordinates. The intensity or power of the radar signals detected by the two radar sensors can be assigned to locations within this volume, describing the radar signal reflection intensity of objects positioned at those locations. The acquired radar signals can be analyzed using different intervals (i.e., bins) defined based on a discretization of the respective common volume. For example, bins can be defined using distance and radial velocity. Each interval or bin within this volume can, for example, have two channels: a power or intensity spectrum determined using first radar signal data from the first radar sensor and a power or intensity spectrum determined using second radar signal data from the second radar sensor. For example, a machine learning module, including an architecture such as a neural network, can be trained and used to regularize the power spectrum. The neural network can, for example, include convolutional neural network (CNN) layers for regularization.

[0032] For example, a machine learning module is used to determine one or more locations of one or more objects. The machine learning module is trained to provide one or more locations of the one or more objects in response to receiving a first spectrum and a second spectrum as input. Matching and comparing the first and second spectra can be performed, for example, by the machine learning module. In the case of n>2 radar sensors, the machine learning module can, for example, be trained to provide one or more locations of one or more objects in response to receiving first to nth spectra as input.

[0033] For example, a machine learning module may include one or more neural networks. These neural networks may include, for example, one or more of the following: feedforward neural networks, convolutional neural networks. Feedforward neural networks, also known as multilayer perceptrons (MLPs), involve information flow through the network in a single direction, from the input layer through one or more hidden layers to the output layer, without any recurrent or feedback connections. Convolutional neural networks are deep learning models specifically designed for image classification, object detection, and image segmentation. Their ability to automatically learn hierarchical representations from raw input data makes them well-suited for analyzing images and other grid-like data structures. A convolutional neural network consists of an input layer, hidden layers, and an output layer. Hidden layers include one or more layers that perform convolutions, i.e., convolutional layers. As the convolutional kernel slides along the input matrix of the layer, the convolution operation generates feature maps, which in turn contribute to the input of the next layer. These one or more convolutional layers may be followed by other layers, such as pooling layers, fully connected layers, and / or normalization layers.

[0034] For example, Fast-ABC can be used as a neural network, as described in Xiaoru Xie et al., “Fast-ABC: A Fast Architecture for Bottleneck-Like Based Convolutional Neural Networks,” 2019 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), pp. 1–6. Fast-ABC provides an accelerator for BLO (Bottleneck-Like Operation) based convolutional networks.

[0035] Neural networks may be able to efficiently learn matching between the spectra at locations with and without objects. These neural networks are trained to fuse the power spectra at different locations in a global coordinate system. Since the idea of ​​stereo radar matching has not been explored in previous research, concepts used in stereo algorithms for optical images can be leveraged to design neural network architectures. Stereo algorithms for optical images can be used to extract depth information from stereo image pairs by matching corresponding points between pairs of images, thereby inferring the 3D structure of objects with higher accuracy.

[0036] For example, a machine learning module includes one or more encoder-decoder blocks with residual layers, which are trained to determine the similarity between the first spectrum and the second spectrum as output in response to receiving a first spectrum and a second spectrum as input.

[0037] Because of noise in radar signal data, the power spectra of two radar sensors may not perfectly match and show some deviations from each other. Therefore, the data can be passed to one or more encoder-decoder blocks with residual layers (i.e., hourglass blocks) to regularize the data, find similarities between channels, and mitigate the effects of noise. These encoder-decoder blocks can also extract information at various levels of detail, from fine-grained details to more global details, which helps improve resolution. The output of the neural network can indicate the location of an object. Since the results from two radars are used, by matching and comparing the first and second spectra of the corresponding first and second radar sensors, the location of one or more objects in terms of azimuth and elevation can be determined with improved accuracy. To reduce processing time and avoid unnecessary computation, the computation can be limited to the detection area of ​​the radar sensor from which the intensity peaks are obtained. Therefore, the computation can be limited to the region containing potential objects.

[0038] For example, a machine learning module to be trained can be provided. A set of training datasets can be provided for training the machine learning module to be trained. For example, each training dataset may include a first training spectrum and a second training spectrum, as well as training specifications for one or more locations of one or more objects. The first training spectrum describes the strength of a first training radar signal as a function of azimuth and elevation. The second training spectrum describes the strength of a second training radar signal as a function of azimuth and elevation. The machine learning module to be trained can be trained to provide, in response to receiving the first and second training spectra of each training dataset as input, one or more locations of one or more objects as output, defined by the training specifications of the training datasets.

[0039] For n>2 radar sensors, the training dataset may, for example, include first to nth training spectra and training specifications for one or more locations of one or more objects. The first to nth training spectra describe the strength of the first to nth training radar signals as a function of azimuth and elevation angles. A machine learning module can be trained to provide, in response to receiving the first to nth training spectra of each training dataset as input, one or more locations of one or more objects as output, defined by the training specifications of the training dataset.

[0040] For example, determining one or more vector velocities of one or more objects. Determining one or more vector velocities involves using one or more positions of the determined one or more objects and one or more radial velocities of the one or more objects determined using first and second radar signal data. The vector velocity of an object refers to the change of its position and radial velocity over time, providing a comprehensive description of the object's motion. Obtaining the total velocity (including its magnitude and orientation) of a detected object can be particularly useful, for example, in an autonomous driving system.

[0041] In another aspect, a computer program is disclosed for detecting one or more objects using at least two radar sensors arranged spaced apart from each other in a stereo radar assembly. The computer program includes program instructions executable by a processor of a computer device to cause the computer device to receive first radar signal data determined using a first radar sensor of the at least two radar sensors. The first radar signal data describes the characteristics of the first radar signal acquired using the first radar sensor. Second radar signal data determined using a second radar sensor of the at least two radar sensors is received. The second radar signal data describes the characteristics of the second radar signal acquired using the second radar sensor. Using the first radar signal data, one or more first combinations of range and radial velocity are determined, for which the first radar signal data includes an intensity peak. Using the second radar signal data, one or more second combinations of range and radial velocity are determined, for which the second radar signal data includes an intensity peak. For one or more of the determined one or more first and second combinations of range and radial velocity, the first radar signal data is used to determine a first spectrum describing the intensity as a function of azimuth and elevation. Furthermore, the second radar signal data is used to determine a second spectrum describing the intensity as a function of azimuth and elevation. Furthermore, the first and second spectra are used to determine one or more positions of one or more objects in terms of azimuth and elevation. This determination includes matching and comparing the first and second spectra.

[0042] The program instructions contained in the computer program can also be executed by the processor of the computer device to cause the computer device to perform any example of the computer implementation method described above, for detecting one or more objects using at least two radar sensors arranged spaced apart from each other in a stereo radar assembly.

[0043] For example, a computer program product for detecting one or more objects using at least two radar sensors arranged spaced apart from each other in a stereo radar assembly includes a computer-readable storage medium having program instructions implemented therewith. The program instructions are executable by a processor of a computer device to cause the computer device to receive first radar signal data determined using a first radar sensor of the at least two radar sensors. The first radar signal data describes the characteristics of the first radar signal acquired using the first radar sensor. Second radar signal data determined using a second radar sensor of the at least two radar sensors is received. The second radar signal data describes the characteristics of the second radar signal acquired using the second radar sensor. Using the first radar signal data, one or more first combinations of range and radial velocity are determined, for which the first radar signal data includes an intensity peak. Using the second radar signal data, one or more second combinations of range and radial velocity are determined, for which the second radar signal data includes an intensity peak. For one or more of the determined one or more first and second combinations of range and radial velocity, the first radar signal data is used to determine a first spectrum describing the intensity as a function of azimuth and elevation. Furthermore, the second radar signal data is used to determine a second spectrum describing the intensity as a function of azimuth and elevation. Furthermore, the first and second spectra are used to determine one or more positions of one or more objects in terms of azimuth and elevation. This determination includes matching and comparing the first and second spectra.

[0044] The program instructions provided by the computer program product can also be executed by the processor of the computer device to cause the computer device to perform any example of the computer implementation method described above, for detecting one or more objects using at least two radar sensors arranged spaced apart from each other in a stereo radar assembly.

[0045] In another aspect, a computer device is disclosed for detecting one or more objects using at least two radar sensors arranged spaced apart from each other in a stereo radar assembly. The computer device includes a processor and a memory storing program instructions executable by the processor. The program instructions executed by the processor cause the computer device to receive first radar signal data determined using a first radar sensor of the at least two radar sensors. The first radar signal data describes the characteristics of the first radar signal acquired using the first radar sensor. Second radar signal data determined using a second radar sensor of the at least two radar sensors is received. The second radar signal data describes the characteristics of the second radar signal acquired using the second radar sensor. Using the first radar signal data, one or more first combinations of range and radial velocity are determined, for which the first radar signal data includes an intensity peak. Using the second radar signal data, one or more second combinations of range and radial velocity are determined, for which the second radar signal data includes an intensity peak. For one or more of the determined one or more first and second combinations of range and radial velocity, the first radar signal data is used to determine a first spectrum describing the intensity as a function of azimuth and elevation. Furthermore, the second radar signal data is used to determine a second spectrum describing the intensity as a function of azimuth and elevation. Furthermore, the first and second spectra are used to determine one or more positions of one or more objects in terms of azimuth and elevation. This determination includes matching and comparing the first and second spectra.

[0046] The processor executing program instructions stored in memory can also cause the computer device to perform any example of the computer implementation method described above, for detecting one or more objects using at least two radar sensors arranged spaced apart from each other in a stereo radar assembly.

[0047] It should be understood that one or more of the foregoing examples can be combined as long as the combined embodiments are not mutually exclusive. Attached Figure Description

[0048] The example is described in more detail below with reference to the accompanying drawings, in which:

[0049] Figure 1 A flowchart illustrating an exemplary method for detecting one or more objects using at least two radar sensors arranged spaced apart from each other is shown;

[0050] Figure 2 A schematic diagram illustrating the use of radial distance and angle relative to two radar sensors arranged spaced apart from each other is shown; and

[0051] Figure 3A block diagram of an exemplary computer device is shown for detecting one or more objects using at least two radar sensors arranged spaced apart from each other. Detailed Implementation

[0052] In the following text, similar elements are indicated by the same reference numerals.

[0053] Figure 1 An exemplary method is shown for detecting one or more objects using at least two radar sensors arranged spaced apart from each other.

[0054] In block 200, a first radar sensor receives first radar signal data. In block 202, a second radar sensor receives second radar signal data. The first and second signal data can be received, for example, as separate datasets. The first and second signal data can also be received, for example, together. These received radar signal data describe characteristics of the radar signal, such as intensity. For example, an intensity peak in the received radar signal data indicates object detection. The radar signal data can describe the position (particularly the range) and radial velocity of the object associated with it. Therefore, the signal intensity indicating object detection can be used to determine the possible location of the object and its radial velocity by analyzing the received radar signal data. The first and second radar sensors can operate, for example, in monostatic and / or bistatic modes. In the case of n>2 radar sensors, the first and second radar sensors can operate, for example, in multistatic modes.

[0055] In block 204, one or more first combinations of range and radial velocity using first radar signal data can be determined. In block 206, one or more second combinations of range and radial velocity using second radial signal data can be determined. The first and second combinations of range and radial velocity can each be, for example, combinations of range intervals and radial velocity intervals according to a predefined distribution of intervals forming two-dimensional bins. The first interval distribution used to determine the first combinations of range and radial velocity can be offset relative to the second interval distribution used to determine the second combinations of range and radial velocity.

[0056] In block 208, for the first and second combinations of range and radial velocity determined in blocks 204 and 206, one or more first spectra using the first radar signal data can be determined. In block 210, for the first and second combinations of range and radial velocity determined in blocks 204 and 206, one or more second spectra using the second radar signal data can be determined. A global coordinate system describing azimuth and elevation angles can be used to determine the first and second spectra. The use of the global coordinate system can include converting first local coordinates assigned to a first local coordinate system of the first radar sensor and second local coordinates assigned to a second local coordinate system of the second radar sensor into global coordinates in the global coordinate system.

[0057] In box 212, the determined first and second spectra are matched and compared to determine one or more positions of one or more objects based on azimuth and elevation angles. For example, the method further includes using a machine learning module to determine one or more positions of one or more objects. The machine learning module can be trained to provide one or more positions of one or more objects in response to receiving the first and second spectra as input. The machine learning module may, for example, include one or more neural networks. The one or more neural networks may, for example, include one or more of the following: feedforward neural networks and / or convolutional neural networks. When determining one or more positions of one or more objects, the vector velocities of the detected objects can also be determined. This determination of the vector velocities of the detected objects can be an important component of autonomous driving systems, which need to guide vehicle movement and assist in making decisions related to path planning, obstacle avoidance, speed control, and overall safety.

[0058] Figure 2 An exemplary detection of an object is illustrated using two radar sensors (i.e., a first radar sensor 310 and a second radar sensor 312) arranged spaced apart from each other. Here, an exemplary detection using a monostatic operating mode is shown. For the object, a first power P1 spectrum 302 is determined using first radar signal data acquired using the first radar sensor 310, and a second P2 spectrum 304 is determined using second radar signal data acquired using the second radar sensor 312. Power spectra 302 and 304 can, for example, describe the radar signal strength as a function of azimuth and elevation angles. Figure 2The azimuth angles θ1 and θ2 shown can be determined, for example, in the local coordinate systems assigned to the respective radar sensors 310 and 312. Using coordinate transformation, the power spectra 302 and 304 can be described in a global coordinate system that describes the azimuth and elevation angles defined relative to the radar sensor assembly 314. Using the global coordinate system may include converting first local coordinates assigned to the first radar sensor 310 in a first local coordinate system and second local coordinates assigned to the second radar sensor 314 in a second local coordinate system into global coordinates in the global coordinate system.

[0059] Figure 3 An exemplary computer device 102 is shown for detecting one or more objects using at least two radar sensors arranged spaced apart from each other. Computer device 102 may be integrated into a vehicle (e.g., an automobile). For example, computer device 102 is intended to represent one or more computer devices that can be distributed. Computer device 102 is shown as including a computing system 104. Computing system 104 is intended to represent one or more computing systems. Computer device 102 is also shown as including an optional hardware interface 106. The hardware interface may enable computing system 104 to control other components, such as sensors, like radar sensors for acquiring signal data of objects, if such other components are present. Computing system 104 is further shown as communicating with an optional user interface 108. User interface 108 may also include, for example, a display device, such as a display device in an automobile. The display may include things such as a two-dimensional computer monitor, a touchscreen, a virtual reality system, and an augmented reality system.

[0060] The computing system 104 is further shown as communicating with memory 110. Memory 110 is intended to represent various types of memory that the computing system 104 can access. In one example, memory 110 is a non-transitory storage medium.

[0061] Memory 110 is shown to contain machine-executable instructions 120. The machine-executable instructions 120 enable the computing system 104 to perform various numerical, stereo processing, and computational tasks. The machine-executable instructions 120 also enable the computing system 104 to control and operate other components, such as radar sensors, via hardware interface 106. Execution of the machine-executable instructions 120 by the computing system 104 can cause the computing system 104 to control the computer device 102 to perform methods for detecting one or more objects, for example, such as... Figure 1As shown in the diagram, memory 110 is also shown to include a first spectrum module 122. Memory 110 is also shown to include a second spectrum module 124. The first spectrum module 122 and the second spectrum module 124 are configured to determine characteristics, such as intensity, of the received radar signal data acquired using the first radar sensor and the second radar sensor. Alternatively, a single spectrum module may be used to determine the characteristics of the received radar signal data. Memory 110 is also shown to include a machine learning module 130. Machine learning module 130 may, for example, include two neural network architectures referred to as stereo processing module 126 and encoder-decoder module 128. Stereo processing module 126 may be configured to extract depth information by matching stereo image pairs. It may involve a stereo algorithm for images that matches corresponding points between pairs of images to estimate disparity, which represents pixel-level depth or 3D positional differences between images. For example, the first and second spectra determined using spectrum modules 122, 124 may be provided to stereo processing module 126, for example, as input in the form of two-dimensional images. The encoder-decoder module 128 can be trained to find similarities between spectra and mitigate the effects of noise that may be included in the received radar signal data. The stereo processing module 126 and the encoder-decoder module 128 of the machine learning module 130 can work together to learn and model complex input-output relationships. The machine learning module 130 can also provide complementary functionality to the stereo processing module 126 and the encoder-decoder module 128, whereby the stereo processing module 126 and the encoder-decoder module 128 provide a structured framework for generating an output that can be post-processed by the machine learning module 130, for example, to optimize and / or improve the overall output. In summary, the machine learning module 130 can be trained to provide one or more locations of one or more objects with improved accuracy in response to receiving first and second spectra as input.

[0062] Although the invention has been illustrated and described in detail in the accompanying drawings and the foregoing description, such illustrations and descriptions should be considered illustrative or exemplary rather than restrictive; the invention is not limited to the disclosed embodiments.

[0063] Those skilled in the art, through studying the accompanying drawings, the disclosure, and the claims, will be able to understand and implement other variations of the disclosed embodiments when practicing the claimed invention. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite articles "a" or "an" do not exclude multiple. The mere fact that certain measures are recited in mutually different dependent claims does not imply that combinations of these measures cannot be advantageously used. Any reference numerals in the claims should not be construed as limiting the scope.

[0064] A single processor or other unit may perform the functions of several claims. A computer program may be stored / distributed on a suitable medium, such as an optical storage medium or solid-state medium provided with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems.

[0065] As those skilled in the art will understand, aspects of the present invention can be implemented as apparatus, method, or computer program product. Therefore, aspects of the present invention can take the form of a completely hardware embodiment, a completely software embodiment (including firmware, resident software, microcode, etc.), or an embodiment combining software and hardware aspects, all of which are generally referred to herein as “circuit,” “module,” or “system.” Furthermore, aspects of the present invention can take the form of a computer program product implemented on one or more computer-readable media having computer-executable code implemented thereon.

[0066] Any combination of one or more computer-readable media can be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. As used herein, "computer-readable storage medium" encompasses any tangible storage medium that can store instructions executable by a processor or computing system of a computing device. A computer-readable storage medium can be referred to as a computer-readable non-transitory storage medium. A computer-readable storage medium can also be referred to as a tangible computer-readable medium. In some embodiments, a computer-readable storage medium may also be able to store data accessible by a computing system of a computing device. Examples of computer-readable storage media include, but are not limited to: floppy disks, magnetic hard disk drives, solid-state drives, flash memory, USB thumb drives, random access memory (RAM), read-only memory (ROM), optical discs, magneto-optical discs, and register files of computing systems. Examples of optical discs include compact discs (CDs) and digital universal discs (DVDs), such as CD-ROMs, CD-RWs, CD-Rs, DVD-ROMs, DVD-RWs, or DVD-R discs. The term computer-readable storage medium also refers to various types of recording media that can be accessed by a computer device via a network or communication link. For example, data can be retrieved via a modem, via the Internet, or via a local area network. Computer-executable code implemented on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, fiber optic cable, RF, or any suitable combination thereof.

[0067] Computer-readable signal media may include propagated data signals having computer-executable code implemented therein, for example, in baseband or as a portion of a carrier wave. Such propagated signals may take any of a variety of forms, including but not limited to electromagnetic, optical, or any suitable combination thereof. A computer-readable signal medium may be any computer-readable medium that is not a computer-readable storage medium but can communicate, propagate, or transmit programs for use by or in connection with an instruction execution system, apparatus, or device.

[0068] "Computer memory" or "memory" is an example of a computer-readable storage medium. Computer memory is any memory that can be directly accessed by a computing system. "Computer storage device" or "storage apparatus" is another example of a computer-readable storage medium. A computer storage apparatus is any non-volatile computer-readable storage medium. In some embodiments, a computer storage apparatus may also be computer memory, and vice versa.

[0069] As used herein, "computing system" encompasses electronic components capable of executing programs or machine-executable instructions or computer-executable code. References to computing systems, including examples of "computing system," should be interpreted as potentially including more than one computing system or processing core. A computing system can be, for example, a multi-core processor. A computing system can also refer to a collection of computing systems within a single computer system or distributed across multiple computer systems. The term computing system should also be interpreted as potentially referring to a collection or network of computing devices, each including a processor or computing system. Machine-executable code or instructions can be executed by multiple computing systems or processors, which may be within the same computing device or even distributed across multiple computing devices.

[0070] Machine-executable instructions or computer-executable code may include instructions or programs that instruct a processor or other computing system to perform aspects of the invention. Computer-executable code for performing operations toward aspects of the invention may be written in any combination of one or more programming languages ​​and compiled into machine-executable instructions, including object-oriented programming languages ​​such as Java, Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" programming language or similar programming languages. In some cases, the computer-executable code may be in the form of a high-level language or a pre-compiled form and may be used with an interpreter that generates machine-executable instructions on the spot. In other cases, the machine-executable instructions or computer-executable code may be in the form of programming for programmable gate arrays.

[0071] Computer executable code can reside entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0072] Various aspects of the present invention have been described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block or portion of a block in a flowchart, illustration, and / or block diagram can be implemented, where applicable, by computer program instructions in the form of computer-executable code. It should also be understood that combinations of blocks in different flowcharts, illustrations, and / or block diagrams can be combined without mutual exclusion. These computer program instructions can be provided to a computing system of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, executable via the computing system of the computer or other programmable data processing apparatus, create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0073] These machine-executable instructions or computer program instructions may also be stored in a computer-readable medium that may instruct a computer, other programmable data processing apparatus or other device to function in a particular manner, such that the instructions stored in the computer-readable medium produce an article of writing comprising instructions that implement the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0074] Machine-executable instructions or computer program instructions may also be loaded onto a computer, other programmable data processing apparatus or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide a process for implementing the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0075] As used herein, a "user interface" is an interface that allows a user or operator to interact with a computer or computer system. A "user interface" can also be referred to as a "human-computer interface device." A user interface can provide information or data to and / or receive information or data from an operator. A user interface enables the computer to receive input from the operator and to provide output from the computer to the user. In other words, a user interface allows an operator to control or manipulate a computer, and the interface allows the computer to indicate the effects of the operator's control or manipulation. Displaying data or information on a monitor or graphical user interface is an example of providing information to an operator. Receiving data via a keyboard, mouse, trackball, touchpad, pointing stick, graphics tablet, joystick, game controller, webcam, headset, pedal, wired gloves, remote control, and accelerometer are all examples of user interface components that enable the reception of information or data from an operator.

[0076] As used herein, "hardware interface" encompasses an interface that enables a computer system to interact with and / or control external computing devices and / or devices. A hardware interface allows a computing system to send control signals or instructions to external computing devices and / or devices. A hardware interface also enables a computing system to exchange data with external computing devices and / or devices. Examples of hardware interfaces include, but are not limited to: Universal Serial Bus (USB), IEEE 1394 port, parallel port, IEEE 1284 port, serial port, RS-232 port, IEEE-488 port, Bluetooth connectivity, wireless LAN connectivity, TCP / IP connectivity, Ethernet connectivity, control voltage interface, MIDI interface, analog input interface, and digital input interface.

[0077] As used herein, “display” or “display device” encompasses an output device or user interface suitable for displaying images or data. A display may output visual, audio, and / or tactile data. Examples of displays include, but are not limited to: computer monitors, television screens, touchscreens, tactile electronic displays, and Braille screens.

[0078] Cathode ray tubes (CRTs), storage tubes, bistable displays, electronic paper, vector displays, flat panel displays, vacuum fluorescent displays (VFs), light-emitting diode (LED) displays, electroluminescent displays (ELDs), plasma display panels (PDPs), liquid crystal displays (LCDs), organic light-emitting diode (OLED) displays, projectors, and head-mounted displays.

[0079] The term "machine learning" (ML) refers to a computer algorithm used to extract useful information from a training dataset by automatically building probabilistic models (called machine learning modules or models). Machine learning modules can also be called predictive models. Machine learning algorithms build mathematical models based on sample data called "training data" to make predictions or decisions without being explicitly programmed to perform a task. Machine learning modules can be implemented using learning algorithms such as supervised or unsupervised learning. Machine learning modules can be based on various techniques such as clustering, classification, linear regression, reinforcement learning, self-learning, support vector machines, neural networks, etc. Machine learning modules can be, for example, data structures or programs such as neural networks, particularly convolutional neural networks, support vector machines, decision trees, Bayesian networks, etc. Machine learning modules can be adapted (i.e., trained) to predict unmeasured values. Therefore, it is possible to enable a trained machine learning module to predict unmeasured values ​​as outputs based on other known values ​​as inputs.

[0080] The machine learning module to be trained can be, for example, an untrained machine learning module, a pre-trained machine learning module, or a partially trained machine learning module. The machine learning module being trained can be an untrained machine learning module, trained from scratch. Alternatively, the machine learning module being trained can be a pre-trained or partially trained machine learning module. Typically, it may not be necessary to start with an untrained machine learning module, for example, in deep learning. For example, it can start with a pre-trained or partially trained machine learning module. The pre-trained or partially trained machine learning module may have already been pre-trained or partially trained for the same or similar tasks. Using a pre-trained or partially trained machine learning module can, for example, make it possible to train the trained machine learning module to be trained faster, i.e., training can converge faster. For example, transfer learning can be used to train a pre-trained or partially trained machine learning module. Transfer learning refers to a machine learning process where, when solving a different problem, instead of learning from scratch, one starts from a previously learned pattern. This allows, for example, leveraging previous learning to avoid starting from scratch. A pre-trained machine learning module is a machine learning module previously trained, for example, on a large benchmark dataset to solve a problem similar to the problem to be solved through additional learning. In the case of a pre-trained machine learning module, the previous learning process has already been successfully completed. A partially trained machine learning module is one that has been partially trained, meaning the training process may not yet be complete. Pre-trained or partially trained machine learning modules can, for example, be imported and trained for the purposes disclosed herein.

[0081] List of reference numerals

[0082] 102 Computer Equipment

[0083] 104 Computing System

[0084] 106 hardware interface

[0085] 108 User Interface

[0086] 110 memory

[0087] 120 Machine Executable Instructions

[0088] 122 First Spectrum Module

[0089] 124 Second Spectrum Module

[0090] 126 3D processing module

[0091] 128 encoder-decoder module

[0092] 130 Machine Learning Modules

[0093] 302 First Spectrum

[0094] 304 Second Spectrum

[0095] 306 First Radar Signal Data

[0096] 308 Second Radar Signal Data

[0097] 310 First Radar Sensor

[0098] 312 Second Radar Sensor

[0099] 314 Stereo Radar Components

[0100] d1 Distance between the first radar sensor and the detected object

[0101] d2 Distance between the second radar sensor and the detected object

[0102] θ1 is the azimuth angle measured from the first radar sensor.

[0103] θ2 is the azimuth angle measured from the second radar sensor.

[0104] P1 is the power determined by the first radar sensor.

[0105] P2 power determined by the second radar sensor

Claims

1. A computer-implemented method for detecting one or more objects using at least two radar sensors (310, 312) arranged spaced apart from each other in a stereo radar assembly (314), the method comprising: • Receive using at least two radar sensors (310); The first radar signal data determined by the first radar sensor (310) in 312) describes the characteristics of the first radar signal acquired using the first radar sensor (310); • Receive second radar signal data determined using the second radar sensor (312) of the at least two radar sensors (310; 312), the second radar signal data describing the characteristics of the second radar signal acquired using the second radar sensor (312); • The first radar signal data is used to determine one or more first combinations of range and radial velocity, wherein the first radar signal data includes intensity peaks for the one or more first combinations; • The second radar signal data is used to determine one or more second combinations of range and radial velocity, wherein the second radar signal data includes intensity peaks for the one or more second combinations; • For one or more of the determined first and second combinations of distance and radial velocity: For a given combination of range and radial velocity, the first radar signal data is used to determine a first spectrum (302) describing the intensity as a function of azimuth and elevation, and the second radar signal data is used to determine a second spectrum (304) describing the intensity as a function of the azimuth and elevation. The first and second spectra (302, 304) are used to determine one or more positions of one or more objects in terms of the azimuth and the elevation angle, the determination including matching and comparing the first and second spectra (302, 304).

2. The method of claim 1, wherein the presence of the one or more objects is determined in response to determining the matching positions of intensity peaks in the first and second spectra (302, 304) in terms of azimuth and elevation.

3. The method according to any one of the preceding claims, wherein the first combination and the second combination of distance and radial velocity comprise a combination of distance intervals and radial velocity intervals distributed according to a predefined interval, wherein, The first interval distribution used to determine the first combination of distance and radial velocity is shifted relative to the second interval distribution used to determine the second combination of distance and radial velocity.

4. The method according to any one of the preceding claims further includes determining the first and second spectra (302; 304) using a global coordinate system describing the azimuth and elevation angles, wherein the use of the global coordinate system includes transforming first local coordinates of a first local coordinate system assigned to the first radar sensor (310) and second local coordinates of a second local coordinate system assigned to the second radar sensor (312) into global coordinates of the global coordinate system.

5. The method according to any one of the preceding claims further includes using a machine learning module (130) to determine one or more locations of the one or more objects, the machine learning module (130) being trained to provide one or more locations of the one or more objects in response to receiving the first and second spectra (302; 304) as input.

6. The method according to claim 5, wherein the machine learning module (130) comprises one or more neural networks, the one or more neural networks including, for example, one or more of the following: feedforward neural network, convolutional neural network.

7. The method of claim 6, wherein the machine learning module (130) includes one or more encoder-decoder blocks (128) having residual layers, the residual layers being trained to determine, in response to receiving the first and second spectra (302; 304) as input, the similarity between the first and second spectra as output.

8. The method according to any one of the preceding claims further includes determining one or more vector velocities of the one or more objects, the determination of the one or more vector velocities comprising using one or more positions of the one or more objects determined and one or more radial velocities of the one or more objects determined using the first radar signal data and the second radar signal data.

9. A computer program for detecting one or more objects using at least two radar sensors (310; 312) arranged spaced apart from each other in a stereo radar assembly (314), the computer program comprising program instructions executable by a processor of a computer device (102) to cause the computer device (102): • Receive first radar signal data determined using the first radar sensor (310) of the at least two radar sensors (310; 312), the first radar signal data describing the characteristics of the first radar signal acquired using the first radar sensor (310); • Receive second radar signal data determined using the second radar sensor (312) of the at least two radar sensors (310; 312), the second radar signal data describing the characteristics of the second radar signal acquired using the second radar sensor (312); • The first radar signal data is used to determine one or more first combinations of range and radial velocity, wherein the first radar signal data includes intensity peaks for the one or more first combinations; • The second radar signal data is used to determine one or more second combinations of range and radial velocity, wherein the second radar signal data includes intensity peaks for the one or more second combinations; • For one or more of the determined first and second combinations of distance and radial velocity: For a given combination of range and radial velocity, the first radar signal data is used to determine a first spectrum (302) describing the intensity as a function of azimuth and elevation, and the second radar signal data is used to determine a second spectrum (304) describing the intensity as a function of the azimuth and elevation. The first and second spectra (302, 304) are used to determine one or more positions of one or more objects in terms of the azimuth and the elevation angle, the determination including matching and comparing the first and second spectra (302, 304).

10. A computer device (102) for detecting one or more objects using at least two radar sensors (310; 312) arranged spaced apart from each other in a stereo radar assembly (314), the computer device including a processor and a memory storing program instructions executable by the processor, wherein the program instructions are executed by the processor to cause the computer device (102): • Receive first radar signal data determined using the first radar sensor (310) of the at least two radar sensors (310; 312), the first radar signal data describing the characteristics of the first radar signal acquired using the first radar sensor (310); • Receive second radar signal data determined using the second radar sensor (312) of the at least two radar sensors (310; 312), the second radar signal data describing the characteristics of the second radar signal acquired using the second radar sensor (312); • The first radar signal data is used to determine one or more first combinations of range and radial velocity, wherein the first radar signal data includes intensity peaks for the one or more first combinations; • The second radar signal data is used to determine one or more second combinations of range and radial velocity, wherein the second radar signal data includes intensity peaks for the one or more second combinations; • For one or more of the determined first and second combinations of distance and radial velocity: For a given combination of range and radial velocity, the first radar signal data is used to determine a first spectrum (302) describing the intensity as a function of azimuth and elevation, and the second radar signal data is used to determine a second spectrum (304) describing the intensity as a function of the azimuth and elevation. The first and second spectra (302; 304) are used to determine the position of one or more objects in terms of the azimuth and the elevation angle, the determination including matching and comparing the first and second spectra (302; 304).

Images