Method, system and equipment for removing ground object interference of ground penetrating radar and medium

Abstract

The invention belongs to the related technical field of ground penetrating radars, and provides a ground penetrating radar ground feature interference removal method, system, equipment and medium in order to solve the problem that ground feature interference of an existing ground penetrating radar cannot be completely removed. Extracting ground object reflection energy cluster components of the data after migration based on the generative adversarial network, performing demigration processing on the ground object reflection energy cluster components by adopting a Stolt demigration method under the air electromagnetic wave speed, finally removing ground object reflection from the original ground penetrating radar data, and combining Stolt migration and Stolt demigration technologies to obtain the ground penetrating radar data. And surface object reflection interference in the detection process of the ground penetrating radar is removed by combining the generative adversarial network, so that effective full-automatic removal of surface object reflection is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field related to ground penetrating radar, and in particular relates to a method, system, device and medium for removing ground object interference of ground penetrating radar. Background Art

[0002] The statements in this part merely provide background technical information related to the present invention and do not necessarily constitute prior art.

[0003] Ground Penetrating Radar (GPR) is a radar system used for underground exploration. Different from the radar designed to detect dynamic targets in the air under static conditions, the function of GPR is to transmit electromagnetic waves to the ground during movement and capture the reflections of underground targets. Compared with the underground environment, the dielectric constant difference between the air environment and surrounding objects is often more obvious, and the attenuation of electromagnetic waves in the air is also weaker. Therefore, objects above the ground, including trees, lamp posts, vehicles, buildings, transmission lines, etc., will cause significant reflections or diffractions, resulting in ground object interference noise. This phenomenon is particularly serious in low-frequency and non-shielded antenna configurations. In addition, the air-ground coupling of some GPR devices will cause a considerable amount of energy to radiate into the air, resulting in serious ground object reflection interference in vehicle-mounted or drone GPR. Ground object reflections are very similar in form to the reflections of underground targets. For inexperienced engineers, there are often many misjudgment situations. At the same time, the ground object reflections with strong energy cover up the true reflection information underground. Therefore, it is crucial to effectively suppress ground object reflections.

[0004] Currently, there is no mature automated technology to effectively suppress and remove the phenomenon of ground object reflections. All existing methods require manual discrimination and adjustment and cannot achieve a fully automated process. At the same time, existing methods mostly target the removal of interference from numerical simulation data, while there are significant differences between real GPR data and numerically simulated GPR data. In field acquisitions, the acquisition intervals of GPR are usually not strictly consistent, so ground object reflection events are not strictly hyperbolic reflections. Due to the limitations of extraction methods, most existing removal methods usually cannot obtain ideal results on field data. Summary of the Invention

[0005] To overcome the deficiencies of the above-mentioned prior art, the present invention provides a method, system, device and medium for removing ground object interference of ground penetrating radar, which can effectively remove the reflection interference of surface objects during the detection process of ground penetrating radar.

[0006] To achieve the above object, the present invention adopts the following technical solutions: In the first aspect, the present invention provides a method for removing ground object interference of ground penetrating radar, including: Obtain ground penetrating radar data, and perform migration processing on the ground penetrating radar data at the air electromagnetic wave velocity using the Stolt migration method to obtain a focused ground object reflection energy cluster; For the focused ground object reflection energy cluster, use the trained generative adversarial network to extract the ground object reflection energy cluster components, and perform inverse migration processing on the ground object reflection energy cluster components at the air electromagnetic wave velocity using the Stolt inverse migration method to obtain the ground object reflection data after inverse migration; Subtract the ground object reflection data after inverse migration from the ground penetrating radar data to obtain the ground penetrating radar data without ground object reflection.

[0007] In a second aspect, the present invention provides a ground penetrating radar ground object interference removal system, including: A migration processing module, which is configured to: obtain ground penetrating radar data, and perform migration processing on the ground penetrating radar data at the air electromagnetic wave velocity using the Stolt migration method to obtain the migrated data; An inverse migration processing module, which is configured to: for the focused ground object reflection energy cluster, use the trained generative adversarial network to extract the ground object reflection energy cluster components, and perform inverse migration processing on the ground object reflection energy cluster components at the air electromagnetic wave velocity using the Stolt inverse migration method to obtain the ground object reflection data after inverse migration; A ground object interference module, which is configured to: subtract the ground object reflection data after inverse migration from the ground penetrating radar data to obtain the ground penetrating radar data without ground object reflection.

[0008] In a third aspect, the present invention provides an electronic device, including a memory and a processor, as well as computer instructions stored on the memory and running on the processor. When the computer instructions are run by the processor, the method described in the first aspect is completed.

[0009] In a fourth aspect, the present invention provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the method described in the first aspect is completed.

[0010] The above one or more technical solutions have the following beneficial effects: The present invention performs migration processing on the collected ground penetrating radar data at the air electromagnetic wave velocity using the Stolt migration method, then extracts the ground object reflection energy cluster components of the migrated data based on the generative adversarial network, and then performs inverse migration processing on the ground object reflection energy cluster components at the air electromagnetic wave velocity using the Stolt inverse migration method. Finally, the ground object reflection is removed from the original ground penetrating radar data. Combining the Stolt migration and Stolt inverse migration techniques, as well as the joint generative adversarial network, the reflection interference of surface objects during the ground penetrating radar detection process is removed, realizing the effective full automation of removing ground object reflection.

[0011] Advantages of additional aspects of the present invention will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The accompanying drawings forming a part of this specification are used to provide a further understanding of the present invention. The schematic embodiments and descriptions thereof of the present invention are used to explain the present invention and do not unduly limit the present invention.

[0013] Figure 1 It is a flowchart of a method for removing ground object interference of a ground penetrating radar in the first embodiment of the present invention; Figure 2 It is a schematic diagram of the original radar data collected in the first embodiment of the present invention and after data verification and removal of zero drift; Figure 3 It is a schematic diagram of the processing result after Stolt migration of the original data at the electromagnetic wave velocity in the air environment in the first embodiment of the present invention; Figure 4 It is a schematic diagram of the processing result of the migrated data by using an energy cluster extractor of a generative adversarial network in the first embodiment of the present invention; Figure 5 In the first embodiment of the present invention, at the electromagnetic wave velocity in the air environment, for Figure 4 It is a schematic diagram of the processing result after Stolt inverse migration of the extracted data; Figure 6 It is a schematic diagram of the processing result of separating ground object reflections in the original data in the first embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0015] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention.

[0016] In the case of no conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0017] Embodiment 1 This embodiment discloses a method for removing ground object interference of a ground penetrating radar, including: Obtaining ground penetrating radar data, and performing migration processing on the ground penetrating radar data at the electromagnetic wave velocity in the air by using the Stolt migration method to obtain migrated data; For the reflected energy cluster of the focused ground object, the trained generative adversarial network is used to extract the component of the ground object reflected energy cluster, and the Stolt de-migration method is used to perform de-migration processing on the component of the ground object reflected energy cluster at the air electromagnetic wave velocity to obtain the de-migrated ground object reflection data; Subtract the de-migrated ground object reflection data from the ground penetrating radar data to obtain the ground penetrating radar data without ground object reflection.

[0018] This embodiment combines the Stolt migration and Stolt de-migration techniques, as well as the joint generative adversarial network to remove the surface object reflection interference in the ground penetrating radar detection process, and can achieve effective full automation removal.

[0019] The following combines Figure 1 A method for removing ground object interference of a ground penetrating radar proposed in this embodiment is described in detail: Step 1: Obtain the ground penetrating radar data, and perform migration processing on the ground penetrating radar data at the air electromagnetic wave velocity using the Stolt migration method to obtain the migrated data.

[0020] This embodiment has certain requirements for the acquisition process of the ground penetrating radar data. The Stolt migration method adopted in this embodiment is based on the basic assumption of the explosive reflection interface, and the spacing between the transmitting and receiving antennas should be zero in the assumption. Therefore, when collecting the ground penetrating radar data, the transmitting and receiving antennas should be self-exciting and self-receiving antennas, or a ground penetrating radar antenna layout mode with a small transmitting and receiving distance should be adopted. In principle, for high-frequency antennas with a center frequency greater than 100 MHz, the transmitting and receiving distance should not be higher than 0.5 m, and for low-frequency antennas with a center frequency less than or equal to 100 MHz, the transmitting and receiving distance should not be higher than 1 m.

[0021] It should be noted that in the process of processing the collected ground penetrating radar data, the correct travel time and measurement point spacing are crucial for accurately migrating the ground object reflection in the ground penetrating radar data. Therefore, data verification is required in the preprocessing of the ground penetrating radar data to ensure that all travel time information and measurement point spacing are correct for the subsequent correct migration process.

[0022] It is also necessary to preprocess the collected ground penetrating radar data to ensure the data quality. Specifically, the collected ground penetrating radar data is processed for zero-point drift removal, DC removal, mean value removal, deconvolution processing, and signal enhancement processing to ensure that the signal has sufficient intensity amplitude.

[0023] As Figure 2 shown, it is a schematic diagram of the ground penetrating radar data after data verification, including two parts: ground object reflection and underground reflection.

[0024] High-precision migration is the key to ensuring the accurate restoration and removal of ground object reflections. In Stolt migration and Stolt inverse migration, the interpolation deviation in different spaces is an important cause of errors. In Stolt migration and Stolt inverse migration, during the transformation from the time-depth domain to the frequency-wavenumber domain, sine function interpolation is used for spatial mapping to avoid the appearance of false event axes and artifacts and improve the smoothness of imaging.

[0025] For the processed ground penetrating radar data, Stolt migration is performed at the electromagnetic wave velocity in air (0.3 m / ns) to relocate diffracted waves, focus reflected waves, and effectively separate them from other underground reflection signals.

[0026] Specifically, during the Stolt migration process, the ground penetrating radar data with ground object reflection events can be expressed as:

[0027] where, represents the 2D observation field data of the ground penetrating radar collected by the horizontal survey line; ASD and represent the observed ground object reflection event and the true underground reflection, respectively.

[0028] Therefore, the Stolt migration process can be expressed as:

[0029] where,

[0030] where c is the speed of light in air, 、 represent the wavenumbers in two directions, respectively; is the Fourier transform result of the original observation field , . is the Stolt migration operator.

[0031] In this process, this embodiment uses a velocity of 0.3 m / ns for migration. Therefore, the image field consists of two parts, where, represents the image of the object on the true surface; is the migration result of the underground reflection, which is considered clutter here.

[0032] As Figure 3 shows, for the processing result after Stolt migration at the electromagnetic wave velocity in the air environment, the ground object reflection is relocated and focused.

[0033] Step 2: For the focused ground object reflection energy cluster, the trained generative adversarial network is used to extract the ground object reflection energy cluster components, and the Stolt de-migration method is used to de-migrate the ground object reflection energy cluster components at the air electromagnetic wave speed to obtain the de-migrated ground object reflection data.

[0034] To extract ground object reflection energy clusters using a generative adversarial network, a deep learning-based ground-penetrating radar (GPR) offset energy cluster extractor must be built. This requires an effective deep learning dataset. This requires a sufficiently large dataset, adequate generalization methods, and data that closely resembles real-world scenarios. Therefore, synthetic data is used as the dataset. The dataset consists of two parts: a subsurface reflection dataset and a ground object reflection dataset. For the subsurface reflection data, ground-penetrating radar data from a long survey line is used. For the ground object reflection data, numerical data simulated using numerical simulations is used.

[0035] Specifically, the construction of the underground reflection dataset includes: (1) Screening of underground reflection datasets.

[0036] Check all data in the subsurface reflection dataset and remove any obsolete or poor quality data. Also, remove any data containing ground object reflection events to ensure the accuracy of the dataset.

[0037] (2) Processing and creation of underground reflection datasets.

[0038] The underground reflection data sets used include long-line ground penetrating radar data ranging from hundreds of meters to several kilometers. The acquired long-line ground penetrating radar data should first be processed as necessary, including data verification, zero drift processing, DC removal processing, mean removal processing, deconvolution processing and signal enhancement processing.

[0039] (3) Expansion of underground reflection dataset.

[0040] Due to the limited data availability of the existing subsurface reflectance dataset, a series of methods were employed to expand it. Two equal-length segments were randomly cropped from the long-line ground-penetrating radar data, with the lengths of the segments randomly set between 50 and 100 meters. These two segments were then overlaid to achieve data augmentation.

[0041] A data enhancement method was introduced to add Gaussian noise with random variance and perform horizontal flipping and phase flipping on the underground reflection data.

[0042] The construction of the ground object reflection dataset is as follows: a random target model randomly distributed in the air background is used for numerical simulation to simulate the B-SCAN data of ground object reflection in a 3D scene.

[0043] The random target is created by adding random Berlin noise to an ellipse with random major and minor axes, thereby establishing a random target model.

[0044] Process the ground object reflection dataset: Crop the simulated ground object reflection data with the same length as the underground reflection data for subsequent combination and synthesis of the dataset.

[0045] In actual ground penetrating radar data, due to uneven movement or uneven point spacing, the ground object reflection is often not a strict hyperbola, which is different from the result of numerical simulation. Therefore, the Stolt migration results of real ground object reflection data usually do not strictly cluster together. To make the simulated ground object reflection data more similar to the field data, some data segments are scaled and adjusted to simulate uneven velocity or uneven point spacing. During this process, the range of each scaled data segment is randomly selected from 2 to 10 m, the scaling factor is randomly selected from 0.75 to 1.25, and the scaling frequency is randomly selected between 5 and 15 times.

[0046] Construct the underground hyperbolic reflection dataset, specifically: Use the random target model randomly distributed in the underground background for numerical simulation to simulate the B-SCAN data of underground reflection in a 3D scene.

[0047] Add a simulated underground reflection hyperbola showing the underground environment velocity to each underground reflection data. This step expands the dataset, prevents chaos in the network, and ensures that all reflection hyperbolas are correctly identified.

[0048] Randomly combine the processed underground reflection dataset, underground hyperbolic reflection data, and simulated random ground object reflection data obtained above. The combination method is to directly superimpose them in amplitude.

[0049] Offset the randomly combined dataset (i.e., the underground reflection dataset, underground hyperbolic reflection data, and simulated random ground object reflection data) at the electromagnetic wave velocity in the air as the input of the generative adversarial network, and offset the corresponding ground object reflection dataset at the electromagnetic wave velocity in the air as the output of the generator in the generative adversarial network, and train the generative adversarial network.

[0050] In this embodiment, a generative adversarial network is used as the ground penetrating radar migration energy cluster extractor. The generative adversarial network includes a generator and a discriminator. The work of the generator is to extract the focused ground object reflection data according to the input randomly combined dataset, making the extracted focused ground object reflection data indistinguishable from the real focused ground object reflection.

[0051] The role of the discriminator is to distinguish between the real ground object reflection focused data and the focused ground object reflection data generated by the generator, and improve the discrimination ability. The generator and the discriminator compete with each other, and the two networks iterate to improve performance.

[0052] The generator uses three dilated convolutions to convert the input data into feature maps of three different scales; then, an ensemble residual convolution block processes the feature map of the smallest scale and fuses it with the feature maps of other scales; the fused feature maps are further processed by the residual convolution block; finally, the offset ground object reflection data can be generated and extracted.

[0053] For the feature fusion part in the generative adversarial network, first calculate the attention scores from one feature map to another and save them; then perform multiple convolution operations on one of the feature maps to obtain two modulation parameters a and b; the trained modulation parameters are embedded into the other feature map.

[0054] Use the training dataset to train the constructed generative adversarial network. Use the Adam optimizer, and the learning rate starts from 0.001 and decays exponentially. The number of iterations during training should not be less than 150,000 times to obtain a trained ground penetrating radar offset energy cluster extractor.

[0055] After offsetting the actually collected ground penetrating radar data, perform normalization and standardization. Scale the data of different sizes proportionally to the data interval from -1 to 1, and use first-order interpolation to interpolate the data to the standard two-dimensional data size of 512×512; use the standardized data as the test set data to input into the generative adversarial network extractor. After processing, the extracted data is obtained; the extracted data is the offset result of the ground penetrating radar data with only ground object reflections at the electromagnetic wave speed in the air.

[0056] For the ground object reflection energy cluster component of the offset data extracted by the ground penetrating radar offset energy cluster extractor, the background value is usually non-zero and there is weak clutter, which may affect subsequent amplitude alignment. Therefore, data elements with absolute values lower than 5% of the maximum absolute value should be set to zero to remove all low-energy background data.

[0057] As Figure 4 shown, using the ground penetrating radar offset energy cluster extractor based on the generative adversarial network to process the offset data, the offset signals of ground object reflections are extracted, and the offset signals of underground reflections are filtered out.

[0058] Before entering the ground penetrating radar offset energy cluster extractor, the offset data is scaled to the size of 512×512. Therefore, it is necessary to reconstruct the scale size of the extracted ground object reflection energy cluster component to ensure the consistency of the de-offset result with the original ground penetrating radar data. The two-dimensional data output by the ground penetrating radar offset energy cluster extractor is reconstructed to the scale size of the original ground penetrating radar data by using first-order interpolation.

[0059] Before entering the ground penetrating radar migration energy cluster extractor, the migrated data is scaled to a size between -1 and 1. Therefore, it is necessary to reconstruct the amplitude energy of the extracted feature reflection energy cluster components to ensure the consistency between the de-migrated result and the original ground penetrating radar data. The energy alignment is achieved by replacing all the extracted feature reflection energy cluster amplitudes with the amplitudes of the original ground penetrating radar data. In this process, the extracted data after background removal serves as a mask for data restoration.

[0060] For the extracted feature reflection energy cluster components, Stolt de-migration is performed at the electromagnetic wave velocity in air, such as 0.3 m / ns, to restore all feature reflections to their original states. In this process, the de-migration process can be expressed as:

[0061] where, represents the frequency of the radar signal, can be replaced by the following equation, represent the wave number in the x direction and the z direction respectively. represents the ASD data after Fourier transform:

[0062] where c is the speed of light in air, 0.3 m / ns, is the Fourier transform result of the focused feature reflection. After de-migration, an observation field with only the original feature reflection events can be obtained.

[0063] High-precision migration and de-migration are the keys to ensuring the accurate restoration of feature reflections. At the same time, during de-migration, it is necessary to expand the data boundaries and fill zeros at the left and right boundaries and the lower boundary to eliminate boundary artifacts and improve the restoration accuracy of de-migration. For the result after de-migration, the de-migrated result is cropped and intercepted using the original data size and data position, and the intercepted result should correspond one-to-one with the result before de-migration in terms of position.

[0064] As Figure 5 shown, at the electromagnetic wave velocity in the air environment, the processing result after Stolt de-migration of the Figure 4 extracted data restores the feature reflection to the shape and amplitude of the original data.

[0065] Step 3: Subtract the de-migrated feature reflection data from the ground penetrating radar data to obtain the ground penetrating radar data without feature reflections.

[0066] Directly subtract the de-migrated feature reflection separation result from the original ground penetrating radar data to initially obtain the data without feature reflections.

[0067] A spike removal filter is set up to smooth the initially obtained data without ground object reflections and remove spikes.

[0068] The spike removal filter includes a moving average filter and a band-pass filter. The moving average filter smooths the data, aiming to eliminate spikes and rough noises in the processing results. For each trace of data, the smoothed data is the average of the three adjacent traces of the original data.

[0069] The band-pass filter smooths the data, also aiming to eliminate high-frequency spikes and rough noises in the processing results. For each trace of data, it is processed separately using the band-pass filter. Since this band-pass filter only targets ultra-high-frequency spike noises and background low-frequency noises, in order not to affect the fully automated processing flow for ground-penetrating radar equipment with different frequencies, its band-pass range is 5 MHz - 800 MHz.

[0070] As Figure 6 shown, in the processed result of separating ground object reflections from the original ground-penetrating radar data, only underground reflection signals remain in the data.

[0071] In this embodiment, for ground-penetrating radar data interfered by ground object echo signals in the field, first, strict data verification is performed on the data, and then the data is migrated at the air electromagnetic wave velocity using the Stolt migration method to obtain a focused ground object reflection energy cluster. The ground-penetrating radar migration energy cluster extractor based on the generative adversarial network receives the migrated data and fully extracts the ground object reflection energy cluster components. Then, the Stolt inverse migration method is used to restore from the energy cluster to the original ground object reflection. Finally, the ground object reflection is separated from the original data. This implementation scheme combines Stolt migration and inverse migration techniques, generative adversarial network techniques, and various filter techniques, and can effectively and automatically remove the interference of surface object reflections during the ground-penetrating radar detection process.

[0072] Embodiment 2 The purpose of this embodiment is to provide a ground-penetrating radar ground object interference removal system, including: A migration processing module, which is configured to: obtain ground-penetrating radar data and perform migration processing on the ground-penetrating radar data at the air electromagnetic wave velocity using the Stolt migration method to obtain the migrated data; An inverse migration processing module, which is configured to: for the focused ground object reflection energy cluster, use the trained generative adversarial network to extract the ground object reflection energy cluster components, and perform inverse migration processing on the ground object reflection energy cluster components at the air electromagnetic wave velocity using the Stolt inverse migration method to obtain the inversely migrated ground object reflection data; A ground object interference module, which is configured to: subtract the inversely migrated ground object reflection data from the ground-penetrating radar data to obtain the ground-penetrating radar data without ground object reflections.

[0073] In more embodiments, there is also provided: An electronic device includes a memory, a processor, and computer instructions stored on the memory and running on the processor. When the computer instructions are run by the processor, the method described in Embodiment 1 is completed. For the sake of brevity, it will not be elaborated here.

[0074] It should be understood that in this embodiment, the processor may be a central processing unit CPU, and the processor may also be other general-purpose processors, digital signal processors DSP, application-specific integrated circuits ASIC, off-the-shelf programmable gate arrays FPGA, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0075] The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A part of the memory may also include a non-volatile random access memory. For example, the memory may also store information about the device type.

[0076] A computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by the processor, the method described in Embodiment 1 is completed.

[0077] The method in Embodiment 1 can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules in the processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory, and the processor reads the information in the memory and combines its hardware to complete the steps of the above method. To avoid repetition, it will not be described in detail here.

[0078] A computer program product includes a computer program. When the computer program is executed by the processor, the method described in Embodiment 1 is implemented.

[0079] The present invention also provides at least one computer program product tangibly stored on a non-transitory computer-readable storage medium. The computer program product includes computer-executable instructions, such as instructions included in program modules, which are executed in a device on a target real or virtual processor to perform the processes / methods described above. Generally, program modules include routines, programs, libraries, objects, classes, components, data structures, etc. that perform specific tasks or implement specific abstract data types. In various embodiments, the functions of program modules can be combined or divided as needed among the program modules. The machine-executable instructions for the program modules can be executed within local or distributed devices. In a distributed device, the program modules can be located in local and remote storage media.

[0080] The computer program code for implementing the method of the present invention can be written in one or more programming languages. This computer program code can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the program code is executed by the computer or other programmable data processing device, the functions / operations specified in the flowchart and / or block diagram are implemented. The program code can be executed entirely on the computer, partially on the computer, as a stand-alone software package, partially on the computer and partially on a remote computer, or entirely on a remote computer or server.

[0081] In the context of the present invention, the computer program code or related data can be carried by any suitable carrier so that a device, apparatus, or processor can perform the various processes and operations described above. Examples of carriers include signals, computer-readable media, and the like. Examples of signals can include electrical, optical, radio, acoustic, or other forms of propagated signals, such as carrier waves, infrared signals, etc. <>

[0082] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in conjunction with this embodiment can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of this application.

[0083] Although the specific implementation manners of the present invention have been described above in conjunction with the accompanying drawings, it is not a limitation on the protection scope of the present invention. Those skilled in the art should understand that, based on the technical solution of the present invention, various modifications or deformations that can be made by those skilled in the art without creative efforts are still within the protection scope of the present invention.

Claims

1. A method for removing ground interference from ground penetrating radar, characterized in that: include: Acquire ground penetrating radar data, and perform migration processing on the ground penetrating radar data using the Stolt migration method at the air electromagnetic wave velocity to obtain migrated data; For the focused ground object reflection energy cluster, the trained generative adversarial network is used to extract the ground object reflection energy cluster components, and the Stolt de-migration method is used to de-migrate the ground object reflection energy cluster components at the air electromagnetic wave speed to obtain the de-migrated ground object reflection data. The ground object reflection data after de-migration is subtracted from the ground penetrating radar data to obtain the ground object reflection-free ground penetrating radar data.

2. The ground object interference removal method for ground penetrating radar according to claim 1, wherein, In Stolt migration and Stolt demigration, spline interpolation or sine function interpolation is used for spatial mapping.

3. The ground object interference removal method for ground penetrating radar according to claim 1, characterized in that The training of the generative adversarial network is specifically as follows: The randomly combined datasets are offset by the speed of electromagnetic waves in the air and used as the input of the generative adversarial network. The corresponding ground object reflection datasets are offset by the speed of electromagnetic waves in the air and used as the output of the generator in the generative adversarial network to train the generative adversarial network. Among them, the randomly combined datasets include underground reflection datasets, underground hyperbolic reflection datasets and simulated random ground object reflection datasets.

4. The ground object interference removal method for ground penetrating radar according to claim 3, characterized in that, The construction of the random object reflection dataset is specifically as follows: Add random Perlin noise to the random major and minor axis ellipse to establish a random target model; Numerical simulation is performed using a random target model randomly distributed in the air background to obtain simulated random ground object reflection data.

5. A method for removing ground object interference of a ground penetrating radar according to claim 1, characterized in that, The Stolt migration method is used to process the ground penetrating radar data at the air electromagnetic wave speed to return the diffraction wave to its original position, focus the reflected wave, and effectively separate it from other underground reflection signals.

6. The ground clutter removal method of a ground penetrating radar according to claim 1, wherein The scale and energy of the energy cluster components reflected by the ground objects are restored, and the restored energy cluster components reflected by the ground objects are de-migrated using the Stolt de-migration method.

7. The method for removing ground interference from ground penetrating radar according to claim 1, wherein: Also includes: The obtained ground penetrating radar data without ground object reflection are smoothed and spikes are removed.

8. A ground penetrating radar ground clutter removal system, characterized in that include: The migration processing module is configured to: acquire ground penetrating radar data, and perform migration processing on the ground penetrating radar data using the Stolt migration method at the air electromagnetic wave velocity to obtain migrated data; A demigration processing module is configured to: extract the ground object reflection energy cluster components using a trained generative adversarial network from the focused ground object reflection energy cluster, perform demigration processing on the ground object reflection energy cluster components using the Stolt demigration method at the air electromagnetic wave velocity, and obtain demigrated ground object reflection data; The ground object interference module is configured to: subtract the ground object reflection data after de-migration from the ground penetrating radar data to obtain ground penetrating radar data without ground object reflection.

9. An electronic device, characterized in that, The method comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the method according to any one of claims 1 to 7 is completed.

10. A computer-readable storage medium, characterized in that, Used to store computer instructions, which, when executed by a processor, complete the method according to any one of claims 1 to 7.

Images