Handheld millimeter wave multi-channel three-dimensional imaging radar system and imaging method

Abstract

The application relates to a handheld millimeter wave multi-channel three-dimensional imaging radar system and an imaging method. The method receives trigger time information of a hardware synchronization controller, radar sampling data of a radar module and position and posture data of a posture sensor; the radar sampling data and the position and posture data are mapped in a time dimension according to the trigger time information to obtain a sampling data set; the sampling data set is filtered through adaptive filtering to obtain a sampling point subset; a three-dimensional image of a target object is obtained through image reconstruction according to the sampling point subset; and the three-dimensional image of the target object is denoised through a dynamic adaptive threshold to obtain a denoised three-dimensional image. The method has the advantages that high synchronization accuracy is realized through a hardware synchronization controller, and the technical effect of improving image quality is realized through adaptive filtering and denoising.

Description

Technical Field

[0001] This invention relates to the field of radar technology applications, and in particular to a handheld millimeter-wave multi-channel three-dimensional imaging radar system and imaging method. Background Technology

[0002] In recent years, millimeter-wave multiple-input multiple-output (MIMO) radar technology has shown great application potential in many fields due to its high integration, wide bandwidth, small size, and low power consumption. Among them, near-field three-dimensional synthetic aperture imaging using compact millimeter-wave radar has become an emerging technology, widely used in scenarios such as security inspection, non-destructive testing, and automotive imaging.

[0003] Existing near-field 3D millimeter-wave imaging systems can be mainly divided into three categories: The first category is systems that employ large-scale regular MIMO arrays. These systems form a large physical aperture through a large number of antennas, enabling real-time imaging using fast algorithms such as range migration (RMA). However, their disadvantages are also significant: the hardware system is bulky, complex in structure, and expensive, and its deployment flexibility is poor, making it unsuitable for portable or handheld applications.

[0004] The second type is the mechanically assisted scanning system. This type of system uses a smaller MIMO radar and assists with devices such as sliding rails and robotic arms to perform equidistant scanning. This approach is less expensive, but it introduces the challenge of synchronizing radar and mechanical movement, requiring precise alignment of mechanical positions with radar operation. Furthermore, the inherent characteristics of the mechanical structure limit its scanning speed and the flexibility of its application scenarios.

[0005] The third category is handheld or random trajectory scanning systems. These systems typically rely on external positioning devices (such as optical tracking systems) to acquire the radar's position and attitude in real time, offering high integration and flexibility. However, existing systems of this type face two critical technical challenges that urgently need to be addressed:

[0006] The synchronization problem between radar data and real-time location: In handheld systems, the timing of radar data acquisition must precisely match the timing of position and attitude information provided by the external positioning system. Traditional software synchronization methods, where a host computer (PC) sends commands and records timestamps, suffer from significant and uncertain delays (typically on the order of 100-200 milliseconds). This temporal mismatch causes radar echo data to be associated with incorrect radar positions, introducing severe phase errors during coherent accumulation imaging, ultimately resulting in severe image defocusing.

[0007] Non-uniform sampling problem: The trajectory of handheld scanning is inherently random and irregular, resulting in unevenly distributed sampling points on the scanning plane—some areas are over-sampled, while others are sparsely sampled. If all acquired data is used indiscriminately in imaging algorithms (such as back projection algorithm BPA), the over-sampled points in the over-dense areas will contribute excessive weight to the final image. This leads to ineffective cancellation of sidelobes and artifacts during coherent accumulation, causing the target signal to be submerged in strong background noise and clutter, resulting in a severe deterioration in image quality and inability to effectively identify the target.

[0008] Currently, no effective solution has been proposed to address the problem of poor image quality caused by inaccurate data position synchronization and non-uniform sampling in handheld radar imaging in related technologies. Summary of the Invention

[0009] The purpose of this invention is to address the shortcomings of existing technologies by providing a handheld millimeter-wave multi-channel three-dimensional imaging radar system and imaging method, thereby solving the technical problems of poor imaging quality caused by inaccurate data position synchronization and non-uniform sampling in handheld radar imaging.

[0010] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0011] This invention provides a handheld millimeter-wave multi-channel 3D imaging radar system, comprising: a radar module, a hardware synchronization controller, an attitude sensor, and a data processing host. The hardware synchronization controller is connected to both the radar module and the data processing host, and is used to receive acquisition commands sent by the data processing host, send trigger signals to the radar module according to the acquisition commands, and return trigger time information to the data processing host based on the trigger signals. The radar module, connected to both the hardware synchronization controller and the data processing host, is used to acquire radar sampling data of the target object based on the trigger signals and return the radar sampling data to the data processing host. The attitude sensor, connected to the data processing host, is used to acquire position and attitude data of the target object and send the position and attitude data to the data processing host. The data processing host receives the trigger time information, the radar sampling data, and the position and attitude data; maps the radar sampling data and the position and attitude data in the time dimension according to the trigger time information to obtain a sample data set; filters the sample data set using adaptive filtering to obtain a subset of sampling points; reconstructs the image based on the subset of sampling points using an image reconstruction algorithm to obtain a 3D image of the target object; and denoises the 3D image of the target object using a dynamic adaptive threshold to obtain a denoised 3D image.

[0012] Optionally, the handheld millimeter-wave multi-channel 3D imaging radar system also includes: a data acquisition card, which comprises: a low-voltage differential signal interface, a data processing module, and a specific communication interface. The data acquisition card is connected to both the radar module and the data processing host. Specifically, the data acquisition card connects to the radar module via a connector to receive radar sampling data through the low-voltage differential signal interface, formats the radar sampling data through the data processing module, and transmits the formatted radar sampling data to the data processing host through the specific communication interface. The data processing host connects to the data acquisition card through the specific communication interface to receive the formatted radar sampling data. The data processing host connects to the attitude sensor via a USB-to-serial adapter to receive position and attitude data. The data processing host reduces the time matching delay between the radar sampling data and the position and attitude data to a threshold based on the trigger time information.

[0013] Optionally, the hardware synchronization controller includes a timer and general-purpose input / output pins, wherein the timer and general-purpose input / output pins send a trigger signal to the radar module according to the acquisition command sent by the data processing host; and return trigger timing information generated based on the trigger signal to the data processing host.

[0014] Furthermore, optionally, the data processing host connects to the radar module via a specific communication interface to configure the radar module's operating parameters; wherein the radar module includes a millimeter-wave radar module.

[0015] This invention provides an imaging method for a handheld millimeter-wave multi-channel 3D imaging radar, applied to a handheld millimeter-wave multi-channel 3D imaging radar system. The method includes: receiving trigger time information from a hardware synchronization controller, radar sampling data from the radar module, and position and attitude data from an attitude sensor; wherein the radar sampling data is sent to a data processing host via a data acquisition card, and the radar module is connected to the data acquisition card; mapping the radar sampling data and position and attitude data along a time dimension based on the trigger time information to obtain a sample data set; filtering the sample data set using adaptive filtering to obtain a subset of sampling points; reconstructing the image based on the subset of sampling points using an image reconstruction algorithm to obtain a 3D image of the target object; and denoising the 3D image of the target object using a dynamic adaptive threshold to obtain a denoised 3D image.

[0016] Optionally, before receiving the trigger time information from the hardware synchronization controller, the radar sampling data from the radar module, and the position and attitude data from the attitude sensor, the method further includes: configuring the operating parameters of the radar module; and sending acquisition commands to the hardware synchronization controller according to a preset period.

[0017] Further, optionally, the radar sampling data and position and attitude data are mapped in the time dimension according to the trigger time information to obtain a sampling data set including: according to the timestamp of the transmission time of each acquisition command in the trigger time information, the position and attitude data corresponding to each frame of radar sampling data at the same transmission time is found in the position and attitude data, resulting in a sampling data set composed of timestamps, each frame of radar sampling data at the same transmission time, and the position and attitude data corresponding to each frame of radar sampling data.

[0018] Optionally, the sampled data set is filtered using adaptive filtering to obtain a subset of sampling points. This includes: calculating the equivalent virtual coverage area on the scanning plane for each sampling point in the sampled data set based on its position and attitude data and the MIMO antenna array parameters of the radar module; wherein the size and orientation of the equivalent virtual coverage area reflect the effective contribution range of the current sampling in space; filtering the sampling points in the sampled data set using a specific strategy, traversing all sampling points in chronological order, selecting the first sampling point into the subset of sampling points, calculating the overlap between the equivalent virtual coverage area of ​​each sampling point after the first sampling point and the union of the equivalent virtual coverage areas of all sampling points selected into the subset of sampling points, discarding sampling points when the overlap is greater than a certain threshold, and selecting sampling points into the subset of sampling points when the overlap is less than a certain threshold.

[0019] Optionally, image reconstruction is performed based on a subset of sampling points using an image reconstruction algorithm to obtain a 3D image of the target object, including: defining a 3D grid in space as the imaging region; for each voxel in the imaging region, traversing all sampling points in the subset of sampling points; calculating the round-trip path distance from the virtual antenna phase center corresponding to each sampling point to the current voxel; performing corresponding phase compensation on the sampled radar echo data based on the round-trip path distance; coherently accumulating the contribution values ​​of all sampling points to the voxel after phase compensation; and determining the accumulated value matrix as the 3D image of the target object after traversing all voxels in the image region.

[0020] Optionally, the denoising of the 3D image of the target object using a dynamic adaptive threshold to obtain the denoised 3D image includes: modeling the 3D image of the target object to obtain a model consisting of the superposition of the effective signal and background noise; setting the background noise amplitude to be much smaller than the target signal, identifying pixels in the model with amplitudes smaller than the target signal as pure noise regions, extracting noise samples from the pure noise regions to obtain a noise sample set; calculating the mean and standard deviation of the noise sample set, and calculating a dynamic adaptive threshold based on the mean and standard deviation using the statistical characteristics of noise; traversing all pixels in the 3D image of the target object, if the pixel amplitude is less than the adaptive threshold, it is determined to be noise and suppressed; if the pixel amplitude is greater than or equal to the adaptive threshold, the pixel amplitude is determined to be an effective signal and retained, and the denoised 3D image is obtained based on the retained pixel amplitude.

[0021] This invention employs the above technical solution, which involves receiving trigger time information from a hardware synchronization controller, radar sampling data from a radar module, and position and attitude data from an attitude sensor; mapping the radar sampling data and position and attitude data along the time dimension based on the trigger time information to obtain a sample data set; filtering the sample data set using adaptive filtering to obtain a subset of sampling points; reconstructing the image using an image reconstruction algorithm based on the subset of sampling points to obtain a 3D image of the target object; and denoising the 3D image of the target object using a dynamic adaptive threshold to obtain a denoised 3D image. Compared with existing technologies, this invention has the following technical effects: achieving high synchronization accuracy through a hardware synchronization controller and improving image quality through adaptive filtering and denoising. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention;

[0023] Figure 2 This is a schematic diagram of the physical composition of a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention;

[0024] Figure 3 This is a system architecture and data flow diagram of a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention;

[0025] Figure 4 This is a conceptual schematic diagram of the adaptive filtering principle in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention;

[0026] Figure 5 This is a schematic diagram of the hardware synchronization controller in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention;

[0027] Figure 6 This is a flowchart illustrating a three-dimensional imaging method in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention.

[0028] Figure 7 This is a schematic diagram of the original imaging effect in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention;

[0029] Figure 8 This is a schematic diagram of the imaging effect after dynamic noise filtering in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention;

[0030] Figure 9 This is a schematic flowchart of an imaging method for a handheld millimeter-wave multi-channel three-dimensional imaging radar according to Embodiment 2 of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0032] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0033] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0034] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units (elements) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or apparatus. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms “multiple” / “several” used in this application refer to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can indicate: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0035] Example 1

[0036] An illustrative embodiment of the present invention, such as Figure 1 As shown, Figure 1 This is a schematic diagram of a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention. The handheld millimeter-wave multi-channel three-dimensional imaging radar system provided in this application includes:

[0037] The system comprises a radar module 10, a hardware synchronization controller 12, an attitude sensor 14, and a data processing host 16. The hardware synchronization controller 12 is connected to both the radar module 10 and the data processing host 16. It receives acquisition commands from the data processing host 16, sends trigger signals to the radar module 10 based on the acquisition commands, and returns trigger time information to the data processing host 16 based on the trigger signals. The radar module 10, connected to both the hardware synchronization controller 12 and the data processing host 16, acquires radar sampling data of the target object based on the trigger signals and returns the radar sampling data to the data processing host 16. The attitude sensor 14 is connected to the data processing host 16. The device 16 is connected to collect the position and attitude data of the target object and sends the position and attitude data to the data processing host 16. The data processing host 16 is used to receive trigger time information, radar sampling data and position and attitude data, and to map the radar sampling data and position and attitude data in the time dimension according to the trigger time information to obtain a sampled data set. The sampled data set is filtered by adaptive filtering to obtain a subset of sampling points. The sampled point subset is used to reconstruct the image by an image reconstruction algorithm to obtain a three-dimensional image of the target object. The three-dimensional image of the target object is denoised by dynamic adaptive thresholding to obtain a denoised three-dimensional image.

[0038] Optionally, in this embodiment of the handheld millimeter-wave multi-channel three-dimensional imaging radar system, a data acquisition card is further included. The data acquisition card includes a low-voltage differential signal interface, a data processing module, and a specific communication interface. The data acquisition card is connected to both the radar module 10 and the data processing host 16. The data acquisition card is connected to the radar module 10 via a connector and is used to receive radar sampling data through the low-voltage differential signal interface, format the radar sampling data through the data processing module, and transmit the formatted radar sampling data to the data processing host 16 through the specific communication interface. The data processing host 16 is connected to the data acquisition card through the specific communication interface and is used to receive the formatted radar sampling data. The data processing host 16 is connected to the attitude sensor 14 via a USB-to-serial port device to receive position and attitude data. The data processing host 16 reduces the time matching delay between the radar sampling data and the position and attitude data to a threshold value based on the trigger time information.

[0039] Optionally, the hardware synchronization controller 12 includes a timer and general-purpose input / output pins, wherein the timer and general-purpose input / output pins send a trigger signal to the radar module 10 according to the acquisition command sent by the data processing host 16; and return trigger timing information generated based on the trigger signal to the data processing host 16.

[0040] Furthermore, optionally, the data processing host 16 is connected to the radar module 10 via a specific communication interface for configuring the operating parameters of the radar module 10; wherein the radar module 10 includes a millimeter-wave radar module.

[0041] Specifically, such as Figure 2 As shown, Figure 2 This is a schematic diagram of the physical composition of a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention; the handheld millimeter-wave multi-channel three-dimensional imaging radar system provided in this application achieves high-precision and high-flexibility three-dimensional imaging through modular hardware design and advanced signal processing algorithms;

[0042] The radar module 10 and data acquisition card in the handheld millimeter-wave multi-channel three-dimensional imaging radar system provided in this application embodiment are... Figure 2 The main components of the handheld scanning probe are integrated with at least two independent millimeter-wave radar transceiver modules, making the handheld scanning probe lightweight and easy for operators to use for scanning.

[0043] The handheld millimeter-wave multi-channel three-dimensional imaging radar system provided in this application embodiment controls the radar module 10 through the hardware synchronization controller 12, so that the data acquisition card acquires raw data from each radar module 10 and ensures that the data acquisition of all channels is strictly synchronized in time.

[0044] The data processing host 16 in the handheld millimeter-wave multi-channel three-dimensional imaging radar system provided in this application embodiment is typically a high-performance industrial computer or workstation, responsible for receiving and processing the massive amounts of data collected, running signal processing algorithms, and finally reconstructing and displaying the three-dimensional image of the target object.

[0045] Figure 3 This is a system architecture and data flow diagram of a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention; as follows: Figure 3 The diagram illustrates the hardware components of the handheld millimeter-wave multi-channel three-dimensional imaging radar system provided in this application embodiment, as well as the connection relationships between these hardware components. The details are as follows:

[0046] The radar module in this embodiment can use Texas Instruments' (TI) xWR1443BOOST evaluation board as the radar front end. The xWR1443BOOST is a standalone frequency modulated continuous wave (FMCW) radar system with three built-in transmit antennas (TX) and four receive antennas (RX), which can form a 12-channel MIMO virtual antenna array. The advantage of the xWR1443BOOST lies in its high integration, including a microcontroller (MCU) and hardware accelerator, which can perform some primary signal processing on-chip, reducing the computational burden on the data processing host 16.

[0047] In a preferred embodiment, the data acquisition card in this application can be a DCA1000 EVM data acquisition card. The xWR1443BOOST is connected to the DCA1000 EVM data acquisition card via a 60-pin connector (i.e., the connector in this application embodiment). The DCA1000 EVM's function is to capture the ADC sampling data (i.e., the radar sampling data in this application embodiment) of the raw intermediate frequency (IF) signal output from the radar front end. The DCA1000 EVM data acquisition card receives the radar sampling data from the xWR1443BOOST through a high-speed LVDS (Low Voltage Differential Signaling) interface, formats the data through the onboard FPGA (i.e., the data processing module in this application embodiment), and finally transmits the radar sampling data to the data processing host 16 in real time through a gigabit Ethernet interface (i.e., the specific communication interface in this application embodiment). The design of connecting the data acquisition card to the data processing host 16 separates high-speed data acquisition from the radar front end, ensuring the stability and bandwidth of data transmission.

[0048] In existing technologies, if multiple modules are triggered solely by software, the startup time of each channel can have an error of hundreds of milliseconds due to factors such as operating system scheduling delays and network transmission jitter. This is fatal for MIMO radar systems that require coherent synthesis. The hardware synchronization controller 12 in this embodiment is a key component to ensure the quality of multi-channel data, and it uses a custom development board based on the STM32F103 microcontroller as the hardware synchronizer. Figure 5 This is a schematic diagram of the hardware synchronization controller in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention; Figure 5 As shown, the hardware synchronization controller 12 in this embodiment of the application ( Figure 5 The STM32 synchronizer (in the middle) generates a periodically stable, steep-edge hardware trigger pulse signal (i.e., the trigger signal in this embodiment) through precise timers and GPIO (general purpose input / output) pins. This trigger signal is simultaneously sent to the xWR1443BOOST via cables of equal length (i.e., dedicated pulse signal lines in this embodiment). When the xWR1443BOOST detects the rising (or falling) edge of this pulse, it immediately triggers the xWR1443BOOST ( Figure 5 The radar module 10 in the system begins acquiring a frame of data. In this way, the start times of acquisition for all radar channels are forcibly aligned, achieving millisecond-level synchronization accuracy, providing a reliable data foundation for subsequent coherent signal processing. After issuing a trigger signal, data is sent to the data processing host 16 according to the trigger signal. Figure 5 The data processing module in the middle returns the trigger time information.

[0049] To achieve non-uniform sampling 3D imaging, the precise position and attitude of the handheld scanning probe in space during each data acquisition is essential. In this embodiment, the attitude sensor 14 can be an FZMotion attitude sensor. The FZMotion attitude sensor integrates a gyroscope, accelerometer, and magnetometer, enabling real-time output of the handheld scanning probe's three-axis angles (pitch, roll, yaw) and acceleration data. These three-axis angles (pitch, roll, yaw) and acceleration data are transmitted to the data processing host 16 via a serial port. The software algorithm on the data processing host 16, combining the three-axis angles (pitch, roll, yaw) and acceleration data, can calculate the motion trajectory of the handheld scanning probe during the scanning process in real time.

[0050] The data processing host 16 can be an industrial computer configured with a high-performance CPU and sufficient memory. This data processing host 16 connects to the DCA1000 EVM via a dedicated Ethernet port (i.e., the specific communication interface in this embodiment) to receive raw MIMO radar data. Simultaneously, the data processing host 16 connects to the FZMotion attitude sensor via a USB-to-serial adapter to receive position and attitude data. The data processing host 16 runs the entire system's control software and core imaging algorithms.

[0051] It should be noted that the components in the handheld millimeter-wave multi-channel three-dimensional imaging radar system provided in this application embodiment are only described as preferred examples to achieve the handheld millimeter-wave multi-channel three-dimensional imaging radar system provided in this application embodiment, and are not specifically limited.

[0052] Figure 6 This is a flowchart illustrating a three-dimensional imaging method in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention. Figure 6 As shown, the imaging method flow of this application embodiment is as follows:

[0053] Step 1: Data Acquisition and Hardware Synchronization. This step aims to acquire a set of spatiotemporally synchronized sampled datasets. The operator sets the radar's operating parameters, such as the start frequency, bandwidth, modulation slope, and frame period of the FMCW signal, on the software of the data processing host 16. These operating parameters are transmitted to the radar module 10 via Ethernet, serial port, or USB port communication.

[0054] Once the system starts operating, the data processing host 16 periodically sends acquisition commands to the hardware synchronization controller 12. The hardware synchronization controller 12 sends hardware trigger pulses (i.e., trigger signals in this embodiment) to the radar module 10 and sends trigger timing information back to the data processing host 16. The data processing host 16 buffers the ADC data stream from the radar module 10 (i.e., radar sampling data in this embodiment) and the position and attitude data from the attitude sensor 14. Using the high-precision trigger timing information provided by the hardware synchronization controller 12, the data processing host 16 can accurately search within the position and attitude data, binding each frame of ADC data to the corresponding position and attitude data at that moment. The motion trajectory calculated from the attitude data can determine the precise coordinates (x, y, z) of each virtual antenna element in three-dimensional space for each virtual antenna element in each frame of ADC data. Due to the handheld scanning, the spatial arrangement of these elements is sparse and non-uniform.

[0055] By sending a trigger signal through a hardware synchronization controller, the data processing host 16 binds the radar sampling data and position and attitude information according to the trigger time information. This reduces the data matching latency from 100-200 milliseconds in traditional software methods to 1-2 milliseconds in hardware communication and processing, thus ensuring the accuracy and reliability of the data foundation for subsequent imaging processing. The output of this step is a set of tuples containing radar data frames, position, attitude, and precise timestamps (i.e., radar sampling data, position and attitude data, and trigger time information in this embodiment).

[0056] Step 2: Adaptive Filtering. This step aims to address the non-uniform sampling problem caused by irregular handheld scanning trajectories. Although the dataset obtained in Step 1 is precisely synchronized in time, it is spatially disordered and unevenly distributed. Directly using this dataset will introduce artifacts. Therefore, this embodiment introduces an adaptive filtering algorithm.

[0057] Figure 4 This is a conceptual schematic diagram of the adaptive filtering principle in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention; see reference. Figure 4 The principle of the adaptive filtering algorithm is as follows: First, for each sampling point in the tuple set, based on its position and attitude data and the MIMO antenna array parameters of the radar module, an equivalent virtual coverage area on the scanning plane is calculated. The equivalent virtual coverage area can be approximated as a rectangle, and the size and orientation of the rectangle reflect the effective contribution range of this sampling in space.

[0058] Then, a greedy strategy is used to filter the sampling points. All sampling points are traversed in chronological order. The first sampling point is selected as the final subset by default. For each subsequent sampling point, the overlap between the equivalent virtual coverage area of ​​each sampling point and the union of the equivalent virtual coverage areas of all selected sampling points is calculated. If this overlap exceeds a pre-set threshold (e.g., 80%), the sampling point is considered spatially redundant because the area it covers has already been fully covered by previous sampling points. Such sampling points (e.g.) Figure 4 (As shown in the dashed box in the image) will be discarded. Conversely, if the overlap is below the threshold, the sampling point is considered to provide new spatial information and is selected into the subset (e.g., ...). Figure 4 (as shown in the solid-line box in the image).

[0059] This process removes a large amount of redundant data from over-sampled areas, while retaining data from sparsely sampled areas. The final output subset of sampled points achieves a more uniform distribution across the entire scanning plane, laying the foundation for high-quality imaging.

[0060] Step 3: Image Reconstruction. The homogenized subset of sampling points after adaptive filtering is input into the image reconstruction algorithm. This embodiment employs the Back Projection Algorithm (BPA). The basic idea of ​​BPA is as follows: First, a three-dimensional grid is defined in space as the imaging region. Then, for each voxel in the imaging region, all sampling points in the sampling point subset are traversed. For each sampling point, the round-trip path distance from the virtual antenna phase center corresponding to the sampling point to the current voxel is calculated, and phase compensation is performed on the sampled radar echo data based on this round-trip path distance. Finally, the contribution values ​​of all sampling points after phase compensation for the voxel are coherently accumulated. After traversing all voxels in the image region, the resulting accumulated value matrix is ​​the three-dimensional image of the target. The calculation process is as follows:

[0061] Assumption It is the reflection function of the target region. Let f be the wavenumber in free space, f be the carrier frequency, and c be the speed of light. Then the corresponding received signal can be expressed as:

[0062] ;

[0063] in, This is the distance history from the launch site to the target area. Let (x', y', z') be the distance history from the receiver to the target area, and (x', y', z') be the three-dimensional spatial coordinates of the target area. t y t , z t (x) represents the three-dimensional spatial coordinates of the transmitting antenna.r， y r， z r (j) represents the three-dimensional spatial coordinates of the receiving antenna, where j is the imaginary unit;

[0064] They can be represented as:

[0065]

[0066] The BPA imaging algorithm can then be expressed as:

[0067]

[0068] Step 4: Dynamic Threshold Noise Reduction. To further improve the quality of the final image, highlight effective target information, and suppress background noise and clutter interference, the signal processing flow in this embodiment also includes a key image post-processing step: a noise suppression method based on adaptive statistical thresholding. The implementation of this method can significantly improve the signal-to-noise ratio and contrast of the image.

[0069] This method first uses radar images Model as an effective signal With background noise The superposition of, that is:

[0070]

[0071] Based on the assumption that the background noise amplitude is much smaller than the target signal, some low-intensity pixels in the image are identified as pure noise regions and extracted as noise samples. Subsequently, the mean and standard deviation of this noise sample set are calculated through statistical analysis to characterize the average intensity and fluctuation range of the noise. Then, based on the statistical characteristics of the noise, a dynamic adaptive threshold T is calculated. The formula for calculating this threshold is:

[0072]

[0073] in, The noise mean. For the noise standard deviation, and This is an adjustable sensitivity coefficient used to balance noise suppression strength and target detail preservation. Finally, all pixels in the image are traversed. If the pixel amplitude is lower than the threshold T, it is determined as noise and suppressed (e.g., set to zero); if it is higher than or equal to the threshold T, it is determined as a valid signal and preserved.

[0074] The above method achieves strong adaptability, automatically adjusting the threshold according to the actual noise level of each image; good statistical robustness, with stable and reliable results based on statistical analysis; and high computational efficiency and simple algorithm, meeting the requirements of rapid system processing.

[0075] Step 5: Image Display. The 3D data matrix generated in Step 4 is visualized, for example, through 2D slicing or 3D rendering, and the image of the target is displayed on the screen of the processing unit. The final generated image is displayed on the host's user interface, allowing the operator to rotate, zoom, and perform other interactive operations to clearly observe the target's internal structure or hidden objects. (Refer to...) Figure 7 and Figure 8 , Figure 7 This is a schematic diagram of the original imaging effect in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention; Figure 8 This is a schematic diagram of the imaging effect after dynamic noise filtering in a handheld millimeter-wave multi-channel three-dimensional imaging radar system according to Embodiment 1 of the present invention. After processing, even under handheld scanning conditions, a clear and distinguishable target image can be obtained, and its effect is far superior to the imaging results without adopting any of the key steps in the present invention.

[0076] In summary, this invention, through the synergistic effect of a hardware synchronizer module and an adaptive filtering algorithm, successfully solves two major technical challenges in handheld millimeter-wave radar imaging, providing a complete and effective technical solution for realizing high-precision and highly flexible portable three-dimensional imaging devices.

[0077] This invention employs the above technical solution, using a hardware synchronization controller connected to both the radar module and the data processing host. This controller receives acquisition commands from the data processing host and sends trigger signals to the radar module based on the acquisition commands. It also returns trigger time information to the data processing host based on the trigger signals. The radar module, connected to both the hardware synchronization controller and the data processing host, acquires radar sampling data of the target object based on the trigger signals and returns the radar sampling data to the data processing host. An attitude sensor, connected to the data processing host, acquires the position and attitude data of the target object and sends the position and attitude data to the data processing host. The data processing host receives the trigger time information, radar sampling data, and position and attitude data. Based on the trigger time information, it maps the radar sampling data and position and attitude data in the time dimension to obtain a sample data set. The sample data set is then filtered using adaptive filtering to obtain a subset of sampling points. An image reconstruction algorithm is used to reconstruct the image based on the subset of sampling points to obtain a three-dimensional image of the target object. Finally, the three-dimensional image of the target object is denoised using a dynamic adaptive threshold to obtain a denoised three-dimensional image. Compared with existing technologies, this invention achieves the following technical effects: high synchronization accuracy is achieved through the hardware synchronization controller, and image quality is improved through adaptive filtering and denoising.

[0078] Example 2

[0079] An illustrative embodiment of the present invention, such as Figure 9As shown, Figure 9 This is a flowchart illustrating an imaging method for a handheld millimeter-wave multi-channel three-dimensional imaging radar according to Embodiment 2 of the present invention. Applied to the handheld millimeter-wave multi-channel three-dimensional imaging radar system in Embodiment 1, the imaging method for a handheld millimeter-wave multi-channel three-dimensional imaging radar provided in this application includes:

[0080] Step S900: Receive trigger time information from the hardware synchronization controller, radar sampling data from the radar module, and position and attitude data from the attitude sensor; wherein, the radar sampling data is sent to the data processing host via a data acquisition card, and the radar module is connected to the data acquisition card;

[0081] Specifically, the imaging method of the handheld millimeter-wave multi-channel three-dimensional imaging radar provided in this application embodiment is applied to the handheld millimeter-wave multi-channel three-dimensional imaging radar system in embodiment 1, particularly to the data processing host side. The data processing host receives trigger time information sent by the hardware synchronization controller, radar sampling data of the radar module acquired by the data acquisition card, and position and attitude data of the attitude sensor. As in the radar module of embodiment 1, the radar module is connected to the data acquisition card, and then the radar sampling data is sent to the data processing host through the data acquisition card. In this application embodiment, the radar module includes a millimeter-wave radar module.

[0082] Optionally, before receiving the trigger time information of the hardware synchronization controller, the radar sampling data of the radar module, and the position and attitude data of the attitude sensor in step S900, the imaging method of the handheld millimeter-wave multi-channel three-dimensional imaging radar provided in this application embodiment further includes: configuring the working parameters of the radar module; and sending acquisition instructions to the hardware synchronization controller according to a preset period.

[0083] Specifically, the radar's operating parameters, such as the start frequency, bandwidth, modulation slope, and frame period of the FMCW signal, are set on the software of the data processing host. These operating parameters are then sent to the radar module via Ethernet, serial port, or USB port communication.

[0084] Once the system starts working, the data processing host periodically (i.e., the preset period in this application embodiment) sends acquisition commands to the hardware synchronization controller.

[0085] Step S902: Based on the trigger time information, the radar sampling data and position and attitude data are mapped in the time dimension to obtain the sampling data set;

[0086] Optionally, in step S902, the radar sampling data and position and attitude data are mapped in the time dimension according to the trigger time information to obtain the sampling data set, which includes: according to the timestamp of the sending time of each acquisition command in the trigger time information, the position and attitude data corresponding to each frame of radar sampling data at the same sending time is found in the position and attitude data, so as to obtain the sampling data set composed of the timestamp, each frame of radar sampling data at the same sending time, and the position and attitude data corresponding to each frame of radar sampling data.

[0087] Specifically, the data processing host caches ADC data (i.e., radar sampling data in this embodiment) and position and attitude data. Utilizing the high-precision trigger timing information provided by the hardware synchronization controller, a precise search is performed within the position and attitude data, binding each frame of ADC data to the corresponding position and attitude data at that moment. The motion trajectory calculated from the attitude data allows for the determination of the precise coordinates (x, y, z) of each virtual antenna element in each frame of ADC data in three-dimensional space. Due to the handheld scanning, the spatial arrangement of these elements is sparse and non-uniform.

[0088] By sending a trigger signal through a hardware synchronization controller, the data processing host binds the radar sampling data and position / attitude information according to the trigger time information. This reduces the data matching latency from 100-200 milliseconds in traditional software methods to 1-2 milliseconds in hardware communication and processing. This ensures the accuracy and reliability of the data foundation for subsequent imaging processing. The output of this step is a set of tuples containing radar data frames, position, attitude, and precise timestamps.

[0089] Step S904: The sampled data set is filtered by adaptive filtering to obtain a subset of sampled points;

[0090] Optionally, the sampled data set is filtered using adaptive filtering to obtain a subset of sampling points. This includes: calculating the equivalent virtual coverage area on the scanning plane for each sampling point in the sampled data set based on its position and attitude data and the MIMO antenna array parameters of the radar module; wherein the size and orientation of the equivalent virtual coverage area reflect the effective contribution range of the current sampling in space; filtering the sampling points in the sampled data set using a specific strategy, traversing all sampling points in chronological order, selecting the first sampling point into the subset of sampling points, calculating the overlap between the equivalent virtual coverage area of ​​each sampling point after the first sampling point and the union of the equivalent virtual coverage areas of all sampling points selected into the subset of sampling points, discarding sampling points when the overlap is greater than a certain threshold, and selecting sampling points into the subset of sampling points when the overlap is less than a certain threshold.

[0091] Specifically, firstly, for each sampling point in the tuple set (i.e., the sampling dataset in this embodiment), an equivalent virtual coverage area on the scanning plane is calculated based on its position and attitude data and the MIMO antenna array parameters of the radar module. The equivalent virtual coverage area can be approximated as a rectangle, the size and orientation of which reflect the effective contribution range of this sampling in space.

[0092] Then, a greedy strategy (i.e., the specific strategy in this embodiment) is used to filter the sampling points. All sampling points are traversed in chronological order. The first sampling point is selected into the final subset by default. For each subsequent sampling point, the overlap between the equivalent virtual coverage area of ​​each sampling point and the union of the equivalent virtual coverage areas of all selected sampling points is calculated. If this overlap exceeds a preset threshold (i.e., the specific threshold in this embodiment) (e.g., 80%), the sampling point is considered spatially redundant because the area it covers has been fully covered by previous sampling points. Such a sampling point is discarded. Conversely, if the overlap is below the threshold, the sampling point is considered to provide new spatial information and is selected into the subset.

[0093] This process removes a large amount of redundant data from over-sampled areas, while retaining data from sparsely sampled areas. The final output subset of sampled points achieves a more uniform distribution across the entire scanning plane, laying the foundation for high-quality imaging.

[0094] Step S906: Based on the subset of sampling points, perform image reconstruction using an image reconstruction algorithm to obtain a three-dimensional image of the target object;

[0095] Optionally, image reconstruction is performed based on a subset of sampling points using an image reconstruction algorithm to obtain a 3D image of the target object, including: defining a 3D grid in space as the imaging region; for each voxel in the imaging region, traversing all sampling points in the subset of sampling points; calculating the round-trip path distance from the virtual antenna phase center corresponding to each sampling point to the current voxel; performing corresponding phase compensation on the sampled radar echo data based on the round-trip path distance; coherently accumulating the contribution values ​​of all sampling points to the voxel after phase compensation; and determining the accumulated value matrix as the 3D image of the target object after traversing all voxels in the image region.

[0096] Specifically, this application employs a back projection algorithm (BPA). The basic idea of ​​BPA is as follows: First, a three-dimensional grid is defined in space as the imaging region. Then, for each voxel in the imaging region, all sampling points in the sampling point subset are traversed. For each sampling point, the round-trip path distance from the virtual antenna phase center corresponding to the sampling point to the current voxel is calculated, and phase compensation is performed on the sampled radar echo data based on this round-trip path distance. Finally, the contribution values ​​of all sampling points after phase compensation for the voxel are coherently accumulated. After traversing all voxels in the imaging region, the resulting accumulated value matrix is ​​the three-dimensional image of the target. The calculation process is as follows:

[0097] Assumption It is the reflection function of the target region. Let f be the wavenumber in free space, f be the carrier frequency, and c be the speed of light. Then the corresponding received signal can be expressed as:

[0098] ;

[0099] in, This is the distance history from the launch site to the target area. Let (x', y', z') be the distance history from the receiver to the target area, and (x', y', z') be the three-dimensional spatial coordinates of the target area. t y t , z t (x) represents the three-dimensional spatial coordinates of the transmitting antenna. r， y r， z r (j) represents the three-dimensional spatial coordinates of the receiving antenna, where j is the imaginary unit;

[0100] They can be represented as:

[0101]

[0102] The BPA imaging algorithm can then be expressed as:

[0103]

[0104] Step S908: The 3D image of the target object is denoised using a dynamic adaptive threshold to obtain the denoised 3D image.

[0105] Optionally, the denoising of the 3D image of the target object using a dynamic adaptive threshold to obtain the denoised 3D image includes: modeling the 3D image of the target object to obtain a model consisting of the superposition of the effective signal and background noise; setting the background noise amplitude to be much smaller than the target signal, identifying pixels in the model with amplitudes smaller than the target signal as pure noise regions, extracting noise samples from the pure noise regions to obtain a noise sample set; calculating the mean and standard deviation of the noise sample set, and calculating a dynamic adaptive threshold based on the mean and standard deviation using the statistical characteristics of noise; traversing all pixels in the 3D image of the target object, if the pixel amplitude is less than the adaptive threshold, it is determined to be noise and suppressed; if the pixel amplitude is greater than or equal to the adaptive threshold, the pixel amplitude is determined to be an effective signal and retained, and the denoised 3D image is obtained based on the retained pixel amplitude.

[0106] Specifically, to further improve the quality of the final image, highlight effective target information, and suppress background noise and clutter interference, the signal processing flow in this embodiment includes a key image post-processing step: a noise suppression method based on adaptive statistical thresholds. The implementation of this method can significantly improve the signal-to-noise ratio and contrast of the image.

[0107] This method first uses radar images Model as an effective signal With background noise The superposition of, that is:

[0108]

[0109] Based on the assumption that the background noise amplitude is much smaller than the target signal, some low-intensity pixels in the image are identified as pure noise regions and extracted as noise samples. Subsequently, the mean and standard deviation of this noise sample set are calculated through statistical analysis to characterize the average intensity and fluctuation range of the noise. Then, based on the statistical characteristics of the noise, a dynamic adaptive threshold T is calculated. The formula for calculating this threshold is:

[0110]

[0111] in, The noise mean. For the noise standard deviation, and This is an adjustable sensitivity coefficient used to balance noise suppression strength and target detail preservation. Finally, all pixels in the image are traversed. If the pixel amplitude is lower than the threshold T, it is determined as noise and suppressed (e.g., set to zero); if it is higher than or equal to the threshold T, it is determined as a valid signal and preserved.

[0112] The above method achieves strong adaptability, automatically adjusting the threshold according to the actual noise level of each image; good statistical robustness, with stable and reliable results based on statistical analysis; and high computational efficiency and simple algorithm, meeting the requirements of rapid system processing.

[0113] Furthermore, the generated 3D data matrix undergoes visualization processing, such as through 2D slicing or 3D rendering, to display the target image on the processing unit's screen. The final generated image is displayed on the host's user interface, allowing the operator to rotate, zoom, and perform other interactive operations to clearly observe the target's internal structure or hidden objects. After processing, even under handheld scanning conditions, a clear and discernible target image can be obtained, with results far superior to imaging results that do not employ any of the key steps in this invention.

[0114] The imaging method of a handheld millimeter-wave multi-channel three-dimensional imaging radar provided in this application can achieve high synchronization accuracy. By introducing a hardware synchronizer module, the time matching error between radar data and position information is reduced from the 100-200 milliseconds level of traditional software methods to an extremely low level of 1-2 milliseconds, fundamentally eliminating the image defocusing problem caused by synchronization delay and providing a foundation for high-quality coherent imaging. Furthermore, it significantly improves image quality. By applying an adaptive filtering algorithm, it effectively suppresses imaging artifacts and noise introduced by non-uniform sampling, enabling the acquisition of clear and highly recognizable target images even under random and irregular scanning trajectories.

[0115] This invention employs the above technical solution, which involves receiving trigger time information from a hardware synchronization controller, radar sampling data from a radar module, and position and attitude data from an attitude sensor; mapping the radar sampling data and position and attitude data along the time dimension based on the trigger time information to obtain a sample data set; filtering the sample data set using adaptive filtering to obtain a subset of sampling points; reconstructing the image using an image reconstruction algorithm based on the subset of sampling points to obtain a 3D image of the target object; and denoising the 3D image of the target object using a dynamic adaptive threshold to obtain a denoised 3D image. Compared with existing technologies, this invention has the following technical effects: achieving high synchronization accuracy through a hardware synchronization controller and improving image quality through adaptive filtering and denoising.

[0116] The above description is merely a preferred embodiment of the present invention and does not limit the implementation and protection scope of the present invention. Those skilled in the art should realize that any equivalent substitutions and obvious changes made based on the description and illustrations of the present invention should be included within the protection scope of the present invention.

Claims

1. A handheld millimeter-wave multi-channel three-dimensional imaging radar system, characterized in that, include: The system consists of a radar module, a hardware synchronization controller, attitude sensors, and a data processing host. The hardware synchronization controller is connected to both the radar module and the data processing host, and is used to receive acquisition instructions sent by the data processing host, send trigger signals to the radar module according to the acquisition instructions, and return trigger time information to the data processing host according to the trigger signals. The radar module is connected to the hardware synchronization controller and the data processing host respectively, and is used to acquire radar sampling data of the target object according to the trigger signal, and return the radar sampling data to the data processing host. The attitude sensor is connected to the data processing host and is used to collect the position and attitude data of the target object and send the position and attitude data to the data processing host. The data processing host is configured to receive the trigger time information, the radar sampling data, and the position and attitude data; map the radar sampling data and the position and attitude data in the time dimension according to the trigger time information to obtain a sampling data set; filter the sampling data set using adaptive filtering to obtain a subset of sampling points; reconstruct an image using an image reconstruction algorithm based on the subset of sampling points to obtain a 3D image of the target object; and denoise the 3D image of the target object using a dynamic adaptive threshold to obtain a denoised 3D image. The step of filtering the sampled data set using adaptive filtering to obtain a subset of sampling points includes: calculating the equivalent virtual coverage area on the scanning plane for each sampling point in the sampled data set based on the position and attitude data of each sampling point and the MIMO antenna array parameters of the radar module; wherein the size and orientation of the equivalent virtual coverage area reflect the effective contribution range of the current sampling in space; filtering the sampling points in the sampled data set using a specific strategy, traversing all sampling points in chronological order, selecting the first sampling point into the subset of sampling points, calculating the overlap degree of the union of the equivalent virtual coverage areas of each sampling point after the first sampling point with the equivalent virtual coverage areas of all sampling points selected into the subset of sampling points, discarding the sampling point when the overlap degree is greater than a specific threshold, and selecting the sampling point into the subset of sampling points when the overlap degree is less than the specific threshold; wherein the specific strategy is a Greedy Strategy; the specific threshold is a pre-set threshold, which is set to 80%; The step of denoising the 3D image of the target object using a dynamic adaptive threshold to obtain the denoised 3D image includes: modeling the 3D image of the target object to obtain a model consisting of the superposition of effective signal and background noise; setting the background noise amplitude to be much smaller than the target signal, identifying pixels in the model with amplitudes smaller than the target signal as pure noise regions, extracting noise samples from the pure noise regions to obtain a noise sample set; calculating the mean and standard deviation of the noise sample set, and calculating a dynamic adaptive threshold based on the mean and standard deviation using the statistical characteristics of noise; traversing all pixels in the 3D image of the target object, if the pixel amplitude is less than the adaptive threshold, it is determined to be noise and suppressed; if the pixel amplitude is greater than or equal to the adaptive threshold, the pixel amplitude is determined to be an effective signal and retained, and the denoised 3D image is obtained based on the retained pixel amplitude.

2. The handheld millimeter-wave multi-channel three-dimensional imaging radar system according to claim 1, characterized in that, The handheld millimeter-wave multi-channel three-dimensional imaging radar system also includes: a data acquisition card, wherein... The data acquisition card includes a low-voltage differential signal interface, a data processing module, and a specific communication interface. The data acquisition card is connected to the radar module and the data processing host respectively. The data acquisition card is connected to the radar module through a connector and is used to receive radar sampling data through the low-voltage differential signal interface, format the radar sampling data through the data processing module, and transmit the formatted radar sampling data to the data processing host through the specific communication interface. The data processing host is connected to the data acquisition card through the specific communication interface to receive the formatted radar sampling data; the data processing host is connected to the attitude sensor through a USB-to-serial device to receive the position and attitude data; the data processing host reduces the time matching delay between the radar sampling data and the position and attitude data to a threshold based on the trigger time information. The specific communication interface is a gigabit Ethernet port.

3. The handheld millimeter-wave multi-channel three-dimensional imaging radar system according to claim 2, characterized in that, The hardware synchronization controller includes a timer and a general-purpose input / output pin. The timer and the general-purpose input / output pin are used to send the trigger signal to the radar module according to the acquisition command sent by the data processing host; and to return the trigger time information generated based on the trigger signal to the data processing host.

4. The handheld millimeter-wave multi-channel three-dimensional imaging radar system according to claim 3, characterized in that, The data processing host is connected to the radar module through the specific communication interface and is used to configure the operating parameters of the radar module; wherein, the radar module includes a millimeter-wave radar module.

5. An imaging method for a handheld millimeter-wave multi-channel three-dimensional imaging radar, characterized in that, The handheld millimeter-wave multi-channel three-dimensional imaging radar system according to any one of claims 1 to 4 includes: The system receives trigger time information from the hardware synchronization controller, radar sampling data from the radar module, and position and attitude data from the attitude sensor; wherein the radar sampling data is sent to the data processing host via a data acquisition card, and the radar module is connected to the data acquisition card; Based on the trigger time information, the radar sampling data and the position and attitude data are mapped in the time dimension to obtain a set of sampling data; The sampled data set is filtered using adaptive filtering to obtain a subset of sampled points; Based on the subset of sampling points, an image reconstruction algorithm is used to reconstruct the image and obtain a three-dimensional image of the target object; The 3D image of the target object is denoised using a dynamic adaptive threshold to obtain the denoised 3D image.

6. The imaging method of the handheld millimeter-wave multi-channel three-dimensional imaging radar according to claim 5, characterized in that, Before receiving the trigger timing information from the hardware synchronization controller, the radar sampling data from the radar module, and the position and attitude data from the attitude sensor, the method further includes: Configure the operating parameters for the radar module; The acquisition command is sent to the hardware synchronization controller according to the preset period.

7. The imaging method of the handheld millimeter-wave multi-channel three-dimensional imaging radar according to claim 5 or 6, characterized in that, The step of mapping the radar sampling data and the position and attitude data in the time dimension based on the trigger time information to obtain the sampling data set includes: Based on the timestamp of the sending time of each acquisition command in the trigger time information, the position and attitude data corresponding to each frame of radar sampling data at the same sending time is searched in the position and attitude data, so as to obtain the sampling data set composed of the timestamp, each frame of radar sampling data at the same sending time, and the position and attitude data corresponding to each frame of radar sampling data.

8. The imaging method of the handheld millimeter-wave multi-channel three-dimensional imaging radar according to claim 7, characterized in that, The step of reconstructing the image based on the subset of sampling points using an image reconstruction algorithm to obtain a 3D image of the target object includes: Define a three-dimensional mesh in space as the imaging area; For each voxel in the imaging region, traverse all sampling points in the subset of sampling points; Calculate the round-trip path distance from the virtual antenna phase center corresponding to each sampling point to the current voxel; The sampled radar echo data are subjected to corresponding phase compensation based on the round-trip path distance; The contribution values ​​of all sampling points to the voxel after phase compensation are coherently accumulated. After traversing all voxels in the imaging region, the resulting accumulated value matrix is ​​determined as the three-dimensional image of the target object.

Images