A method for intelligent repair and reconstruction of a defect image of a ground penetrating radar

Abstract

The application belongs to the technical field of image restoration, and discloses a kind of intelligent restoration and reconstruction processing method for the missing image of ground penetrating radar, including obtaining the radar profile image formed by ground penetrating radar detection, positioning the missing area in the image, and generating missing structure description information according to the spatial adjacency relationship between the missing area boundary and the surrounding effective echo area, and the echo change trend;Call missing structure description information, and perform correlation analysis on the radar echo propagation trajectory of the corresponding position on both sides of the missing area, determine the echo trajectory compensation path passing through the missing area according to the echo continuous propagation relationship;Based on the echo trajectory compensation path, reconstruct the echo information in the missing area according to the corresponding relationship between the missing position and the associated echo trajectory, and generate an initial restoration image;Call initial restoration image, and according to the continuous extension relationship of the reflection boundary of underground medium;Improve the structural consistency and reliability of the reconstruction result of radar profile image.

Description

Technical Field

[0001] This invention relates to the field of image restoration technology, and more specifically, to a method for intelligent restoration and reconstruction of damaged images from ground-penetrating radar. Background Technology

[0002] Ground penetrating radar (GPR), a technology that uses electromagnetic waves to perform non-destructive testing of underground structures, is widely used in fields such as road inspection, underground pipeline detection, and engineering quality assessment. However, in actual testing, due to factors such as the complex distribution of underground media, energy attenuation during electromagnetic wave propagation, and environmental interference during equipment acquisition, the acquired radar profile images are prone to defects such as missing echoes and image breaks, resulting in incomplete representation of some underground reflective interfaces.

[0003] Existing ground-penetrating radar (GPR) image restoration methods typically rely on image interpolation, texture filling, or data fitting to fill in missing areas. These methods primarily rely on the pixel variation patterns around the missing area for restoration, making it difficult to incorporate the spatial continuity during GPR echo propagation. This can easily lead to inconsistencies between the echo variation trend within the restored area and the actual underground reflection structure.

[0004] In addition, the reflective interface of underground media usually has certain extended and continuous characteristics. When repairing based solely on local image information, it is difficult to accurately restore the echo propagation path and the connection relationship of the reflective structure corresponding to the missing area, resulting in structural deviations in the reconstructed radar profile image.

[0005] In view of this, the present invention proposes an intelligent repair and reconstruction method for ground penetrating radar images with defects to solve the above problems. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a method for intelligent repair and reconstruction of defective images from ground-penetrating radar, comprising: S1. Acquire radar profile images generated by ground-penetrating radar detection, locate the defective areas in the images, and generate defective structure description information based on the spatial adjacency relationship between the boundary of the defective area and the surrounding effective echo area, and the echo change trend. S2. Call the description information of the defective structure, perform correlation analysis on the radar echo propagation trajectory at the corresponding positions on both sides of the defective area, and determine the compensation path of the echo trajectory passing through the defective area based on the continuous propagation relationship of the echoes. S3. Based on the echo trajectory compensation path, according to the correspondence between the defect location and the associated echo trajectory, the echo information in the defect area is spatially reconstructed to generate an initial repair image. S4. Call the initial repair image, and based on the continuous extension relationship of the underground medium reflection boundary, perform consistency correction on the reflection interface between the repair area and the non-defective area to generate the target reconstruction image; S5. Obtain the echo continuity relationship in the target reconstructed image, determine the reflection structure association relationship based on the echo continuity relationship, and verify the consistency of the reflection structure association relationship. When the preset consistency condition is met, output the final repair result.

[0007] Preferably, the method for generating the description information of the defective structure includes: The radar profile image is divided into regions to obtain the set of missing regions corresponding to continuous missing regions, and the set of boundary points on the boundary of each missing region is extracted; based on the spatial distance relationship between the set of boundary points and the adjacent effective echo regions, the associated regions corresponding to the missing regions are determined. The system acquires the direction of echo amplitude variation, the trend of continuous echo variation, and the direction of boundary extension within the associated region. It establishes the spatial relationship between the defect area and the surrounding effective echo area, and generates defect structure description information including the defect range, boundary location, and echo extension direction.

[0008] Preferably, the method for determining the echo trajectory compensation path based on the defect structure description information includes: The system retrieves the defect boundary location and echo propagation direction from the defect structure description information to match the effective echo trajectories at corresponding locations on both sides of the defect area; it obtains the echo change relationship between adjacent scanning positions according to the radar scanning direction to determine the degree of continuity between different effective echo trajectories; it filters candidate echo trajectories that meet the continuous echo propagation condition based on the degree of continuity, and determines the candidate echo trajectories as the echo trajectory compensation path that passes through the defect area.

[0009] Preferably, the method for generating the initial repair image includes: The reference echo position corresponding to each location to be repaired within the defect area is determined based on the echo trajectory compensation path; the echo amplitude, direction of change, and propagation distance information corresponding to the reference echo position are obtained; the echo compensation parameters are determined based on the correlation between the reference echo position and the location to be repaired; the echo information of each location to be repaired within the defect area is reconstructed based on the echo compensation parameters, so that the trend of the reconstructed echo change is consistent with the echo trajectory compensation path, and an initial repair image is generated.

[0010] Preferably, the method for determining the echo compensation parameters includes: Obtain the propagation distance, echo change direction, and spatial relationship corresponding to each reference echo position; determine the corresponding distance compensation parameters based on the propagation distance differences between different reference echo positions and the position to be repaired; adjust the echo amplitude corresponding to each reference echo position based on the distance compensation parameters and echo change direction to generate the echo compensation parameters corresponding to the position to be repaired.

[0011] Preferably, the method for generating the target reconstructed image includes: The distribution information of the reflective interface in the initial restoration image is obtained, and the corresponding target extension direction is determined according to the extension direction of the reflective interface in the non-defective region. The direction of the reflective interface in the restoration region is matched with the target extension direction. When there is a directional offset of the reflective interface in the restoration region, the echo change relationship at the corresponding position is adjusted according to the target extension direction. The reflective interface in the restoration region is made to be continuous with the reflective interface in the non-defective region, and the target reconstruction image is generated.

[0012] Preferably, the method for obtaining the echo continuity relationship in the target reconstructed image includes: The target reconstructed image is divided into several detection regions according to a preset spatial range, and the echo change sequence corresponding to each detection region is obtained; the echo continuity relationship between different detection regions is determined based on the echo change difference between adjacent detection regions; the echo propagation direction and continuous change region in the target reconstructed image are determined based on the echo continuity relationship.

[0013] Preferably, the method for determining the correlation of reflection structures based on echo continuity includes: The echo continuity relationship in the target reconstructed image is invoked to trace the echo propagation path in the continuously changing region; the extension direction, connection length and spatial adjacency relationship of the corresponding reflection structure are obtained according to the echo propagation direction; and the correlation relationship of the reflection structure in the target reconstructed image is established based on the extension direction, connection length and spatial adjacency relationship.

[0014] Preferably, the method for verifying the consistency of the correlation between reflection structures includes: The reflection structure association relationship is invoked to obtain the connection state of the reflection structure corresponding to the repaired region; the extension direction, connection length and spatial adjacency relationship of the reflection structure in the repaired region are compared with the corresponding reflection structure in the non-defected region; when the connection state of the reflection structure in the repaired region maintains the correspondence with the reflection structure in the non-defected region, it is determined that the target reconstructed image meets the structural association verification requirements.

[0015] Preferably, the method for outputting the final repair result includes: Based on the reflection structure correlation verification results, the echo change continuous parameters and structure extension correlation parameters corresponding to the target reconstructed image are obtained; the echo change continuous parameters and structure extension correlation parameters are combined and calculated to obtain the reconstruction consistency evaluation results, and the reconstruction consistency evaluation results are compared with the preset consistency conditions; when the reconstruction consistency evaluation results meet the preset consistency conditions, the target reconstructed image is determined as the final repair result and output.

[0016] The technical effects and advantages of the intelligent repair and reconstruction method for damaged images from ground-penetrating radar according to the present invention are as follows: This invention achieves targeted recovery of echo information in the missing area by comprehensively analyzing the spatial correlation between the missing region, the surrounding effective echo region, and the underground reflective structure in a ground-penetrating radar (GPR) image with defects. Compared with existing methods that only fill in gaps based on local grayscale changes or neighboring pixel relationships, this invention can reconstruct the echo change relationship corresponding to the missing area by combining the continuous propagation characteristics of radar echoes in the underground medium. This makes the repaired image information more consistent with the electromagnetic wave propagation and reflection laws during GPR detection. In the process of repairing the defective image, not only is the image information of the missing area itself considered, but also the correlation between the effective echoes on both sides of the missing area is combined to restore the echo propagation path corresponding to the missing location. This avoids the problem of abnormal echo changes caused by the lack of underground structural constraints in traditional image repair processes, ensuring that the missing area maintains a continuous echo change relationship with the surrounding effective detection area.

[0017] Furthermore, by incorporating the extension characteristics of the subsurface medium's reflective interface, the reflective structures in the initial restoration results are adjusted for consistency. This ensures that the reflective interface in the restored area maintains its association with the corresponding structure in the non-defective area, avoiding issues such as positional shifts or inconsistent extension directions of the reflective structures caused by local echo compensation. This improves the accuracy of the reconstructed image in representing subsurface structural information. By establishing the correlation between reflective structures and performing consistency verification, the generated reconstructed image undergoes structural matching judgment. This ensures that the final output not only depends on the echo filling process but also incorporates a comprehensive evaluation based on the connectivity between subsurface reflective structures, thereby improving the reliability and applicability of the damaged image restoration results. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of a method for intelligent repair and reconstruction of damaged images from ground-penetrating radar. Figure 2 This is a schematic diagram of a ground-penetrating radar image defect intelligent repair and reconstruction processing system. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example

[0020] Please see Figure 1 As shown in the figure, this embodiment provides a method for intelligent repair and reconstruction of damaged images from ground-penetrating radar, which specifically includes the following steps: S1. Acquire radar profile images generated by ground-penetrating radar detection, locate the defective areas in the images, and generate defective structure description information based on the spatial adjacency relationship between the boundary of the defective area and the surrounding effective echo area, and the echo change trend. S2. Call the description information of the defective structure, perform correlation analysis on the radar echo propagation trajectory at the corresponding positions on both sides of the defective area, and determine the compensation path of the echo trajectory passing through the defective area based on the continuous propagation relationship of the echoes. S3. Based on the echo trajectory compensation path, according to the correspondence between the defect location and the associated echo trajectory, the echo information in the defect area is spatially reconstructed to generate an initial repair image. S4. Call the initial repair image, and based on the continuous extension relationship of the underground medium reflection boundary, perform consistency correction on the reflection interface between the repair area and the non-defective area to generate the target reconstruction image; S5. Obtain the echo continuity relationship in the target reconstructed image, determine the reflection structure association relationship based on the echo continuity relationship, and verify the consistency of the reflection structure association relationship. When the preset consistency condition is met, output the final repair result.

[0021] Methods for generating description information of defective structures include: The radar profile image is divided into regions to obtain the set of missing regions corresponding to continuous missing regions, and the set of boundary points on the boundary of each missing region is extracted; based on the spatial distance relationship between the set of boundary points and the adjacent effective echo regions, the associated regions corresponding to the missing regions are determined. The system acquires the direction of echo amplitude variation, the trend of continuous echo variation, and the direction of boundary extension within the associated region. It establishes the spatial relationship between the defect area and the surrounding effective echo area, and generates defect structure description information including the defect range, boundary location, and echo extension direction.

[0022] In this embodiment, when dividing the radar profile image into regions, a two-dimensional coordinate relationship is first established according to the radar scanning direction and the electromagnetic wave propagation depth direction. The radar profile image is then divided into multiple continuous detection regions, and the echo information corresponding to the pixels in each detection region is judged. When there are consecutive missing pixels in a certain detection region, and the missing pixels have a continuous spatial connection relationship, the corresponding region is marked as a defect region, and all defect regions are represented as a set to obtain a set of defect regions. Furthermore, the pixels at the boundary positions of each defect region are traversed, and the defect edge positions adjacent to the effective echo region are taken as boundary points to form a corresponding set of boundary points.

[0023] After acquiring the set of boundary points, the associated region is determined based on the spatial distance relationship between the boundary points and the surrounding valid echo areas. Specifically, starting from the boundary point of the defective region, valid echo areas within a preset range are searched along the horizontal and depth directions of the radar image. The spatial distance between the valid echo points and the defective boundary point is calculated, and the valid echo areas that meet the preset distance range are determined as the corresponding associated regions. For example, if the boundary point of the defective region has coordinates (60,5), a search is conducted within a 5-pixel range around this boundary point. If a continuous valid echo area is found within the coordinate range (55,5) to (59,8), and this area is closest to the defective boundary, then this area is determined as the associated region corresponding to the defective region.

[0024] Furthermore, after determining the associated region, echo change information within the associated region is acquired. Specifically, the echo amplitude corresponding to each pixel position within the associated region is read sequentially according to the radar scanning direction, and the amplitude change relationship between adjacent positions is compared to determine the direction of echo amplitude change. Simultaneously, based on the echo change sequence of multiple consecutive positions, it is determined whether the echo amplitude changes continuously along a certain direction, thus obtaining the continuous echo change trend. The boundary extension direction is determined based on the arrangement direction of continuous echo regions within the associated region. For example, if the echo amplitudes of five consecutive positions along the radar scanning direction within a certain associated region are 80, 85, 90, 95, and 100, then it is determined that there is an echo amplitude change direction increasing from 80 to 100 in this region. When adjacent positions maintain an increasing relationship, this change trend is determined as a continuous enhancement trend. If the coordinates of continuous echo points change from (50,10), (55,12), and (60,14), then the boundary extension direction corresponding to this region is determined based on the coordinate change relationship.

[0025] Based on the acquired information such as the extent of the defective region, the set of boundary points, the spatial location of the associated region, the direction of echo amplitude variation, the continuous trend of echo variation, and the direction of boundary extension, these information items are combined to generate defect structure description information. This defect structure description information includes the location range of the defective region, the location of the defective boundary, the location of the associated region, and the direction of echo extension around the defective region, used to represent the spatial correspondence between the defective region and the surrounding valid echo regions.

[0026] Methods for determining the echo trajectory compensation path based on the description information of the defective structure include: The system retrieves the defect boundary location and echo propagation direction from the defect structure description information to match the effective echo trajectories at corresponding locations on both sides of the defect area; it obtains the echo change relationship between adjacent scanning positions according to the radar scanning direction to determine the degree of continuity between different effective echo trajectories; it filters candidate echo trajectories that meet the continuous echo propagation condition based on the degree of continuity, and determines the candidate echo trajectories as the echo trajectory compensation path that passes through the defect area.

[0027] In this embodiment, when calling the defect structure description information, the defect boundary location and echo extension direction information contained therein are first read. Based on the defect boundary location, the effective echo regions corresponding to both sides of the defect area are determined, and the effective echo trajectories located on both sides of the defect area are obtained according to the echo extension direction. The effective echo trajectory on one side of the defect area is used as the starting trajectory, and the effective echo trajectory at the corresponding position on the other side is used as the matching trajectory. Matching is performed according to the positional relationship, extension direction, and echo change trend of the echo trajectories on both sides.

[0028] Furthermore, the echo variation relationship between adjacent scanning positions is obtained according to the radar scanning direction. Specifically, the echo amplitude corresponding to adjacent positions on the same echo trajectory is read sequentially, and the variation between adjacent positions is calculated. Based on the echo variation direction, amplitude variation, and trajectory extension direction of multiple consecutive positions, the degree of continuity between different effective echo trajectories is determined. For example, if the echo amplitude corresponding to a certain effective echo trajectory on the left is 50, 55, and 60 respectively, and the echo amplitude corresponding to the right position is 65, 70, and 75 respectively, and the echoes on both sides gradually increase according to the scanning direction with the same trend, then the two echo trajectories are determined to have a continuous correlation. If the echo amplitude variation of another candidate trajectory is 50, 40, and 30, then its variation direction is inconsistent with the original trajectory, reducing its degree of continuity correlation.

[0029] After determining the degree of continuity correlation between different valid echo trajectories, the degree of continuity correlation of each candidate echo trajectory is compared with preset judgment conditions. Candidate echo trajectories that meet the echo propagation continuity condition are selected, and the selected echo trajectories are used as echo trajectory compensation paths through the defective area. For example, there are three candidate echo trajectories A, B, and C. Trajectory A has echo amplitude changes in the same direction on both sides, with a continuity correlation degree of 0.92; trajectory B has a continuity correlation degree of 0.65; and trajectory C has a continuity correlation degree of 0.40. When the preset continuity correlation judgment value is 0.80, only trajectory A meets the condition, so trajectory A is determined as the echo trajectory compensation path for the corresponding defective area.

[0030] Methods for generating the initial repair image include: The reference echo position corresponding to each location to be repaired within the defect area is determined based on the echo trajectory compensation path; the echo amplitude, direction of change, and propagation distance information corresponding to the reference echo position are obtained; the echo compensation parameters are determined based on the correlation between the reference echo position and the location to be repaired; the echo information of each location to be repaired within the defect area is reconstructed based on the echo compensation parameters, so that the trend of the reconstructed echo change is consistent with the echo trajectory compensation path, and an initial repair image is generated.

[0031] In this embodiment, when determining the reference echo position corresponding to each repairable location within the defective area based on the echo trajectory compensation path, the extended position of the echo trajectory compensation path within the defective area is first obtained, and each pixel position that needs to be restored within the defective area is taken as the repairable location. Based on the spatial coordinate relationship of the repairable location in the echo trajectory compensation path, the corresponding valid echo position is found, and this valid echo position is determined as the reference echo position.

[0032] Furthermore, the echo amplitude, direction of change, and propagation distance information corresponding to the reference echo position are obtained. The echo amplitude represents the intensity of the echo received at the reference position; the direction of change represents the trend of echo amplitude variation between adjacent positions along the radar scanning direction; and the propagation distance represents the distance the radar transmitted signal travels from the detection position to the corresponding reflection position. Based on the spatial relationship between the reference echo position and the position to be repaired, the influence of the reference echo information on the position to be repaired is determined, and corresponding echo compensation parameters are generated. For example, if the echo amplitude at reference echo position A is 80, and the echo amplitude at reference echo position B is 90, and the propagation distances between the two reference positions and the position to be repaired are 10 meters and 15 meters respectively, then the amplitude information corresponding to the two reference echo positions is adjusted according to the difference in propagation distance and the direction of echo change to obtain the echo compensation parameters corresponding to the position to be repaired.

[0033] After obtaining the echo compensation parameters, the echo information within the defect area is reconstructed based on the echo compensation parameters corresponding to each location to be repaired. Specifically, the echo compensation parameters are converted into echo amplitude values ​​corresponding to the locations to be repaired, and each location to be repaired within the defect area is sequentially filled according to the propagation direction in the echo trajectory compensation path, ensuring that the echo variation relationship between adjacent locations corresponds to the echo trajectory compensation path, thus generating an initial repair image. For example, if three locations to be repaired within a defect area are P1, P2, and P3 arranged along the echo trajectory compensation path, and the corresponding calculated echo amplitudes are 70, 75, and 80, then the corresponding echo information is filled into the defect area according to the propagation direction from P1 to P3, forming an initial repair image containing continuous echo variation relationships.

[0034] Methods for determining echo compensation parameters include: Obtain the propagation distance, echo change direction, and spatial relationship corresponding to each reference echo position; determine the corresponding distance compensation parameters based on the propagation distance differences between different reference echo positions and the position to be repaired; adjust the echo amplitude corresponding to each reference echo position based on the distance compensation parameters and echo change direction to generate the echo compensation parameters corresponding to the position to be repaired.

[0035] In this embodiment, when determining the echo compensation parameters, the propagation distance, echo change direction, and spatial position relationship corresponding to each reference echo position are first obtained. The propagation distance is determined based on the distance between the radar scanning position and the reference echo position, representing the positional relationship of the corresponding echo formed after the radar signal propagates. The echo change direction is determined based on the amplitude change relationship of adjacent echo positions around the reference echo position, representing the trend of echo amplitude change along the scanning direction. The spatial position relationship is determined based on the coordinate distance and direction relationship between the reference echo position and the position to be repaired, determining the correspondence between the reference echo information and the position to be repaired. For example, if the coordinates of a position to be repaired are (70,8), and the corresponding two reference echo positions are A(60,8) and B(85,8), then the spatial distance between reference echo position A and the position to be repaired is 10 pixels, and the spatial distance between reference echo position B and the position to be repaired is 15 pixels. Simultaneously, the echo amplitude around reference echo position A increases from 60 to 70, determining that its echo change direction is an increasing direction.

[0036] Furthermore, distance compensation parameters are determined based on the differences in propagation distances between different reference echo positions and the position to be repaired. Specifically, the propagation distances corresponding to each reference echo position are compared with the target propagation distances corresponding to the position to be repaired, and the distance compensation parameters are determined based on the distance difference between the two. When the difference in propagation distances between the reference echo positions and the position to be repaired is large, the corresponding distance compensation parameters are adjusted according to the distance difference. For example, if the propagation distance corresponding to reference echo position A is 20 meters and the propagation distance corresponding to the position to be repaired is 25 meters, there is a 5-meter distance difference between the two, and the corresponding distance compensation parameters are determined based on this distance difference; if the propagation distance corresponding to reference echo position B is 23 meters, which is only 2 meters different from the position to be repaired, then the corresponding distance compensation parameters are different from those for reference echo position A.

[0037] After obtaining the distance compensation parameters, the echo amplitude at the reference echo position is adjusted based on the distance compensation parameters and the echo change direction corresponding to the reference echo position. Specifically, the amplitude change range corresponding to the reference echo position is corrected according to the distance compensation parameters, and the echo amplitude corresponding to the position to be repaired is determined by combining the echo change direction, thus generating the echo compensation parameters for the position to be repaired. For example, if the echo amplitude at reference echo position A is 80, increases along the scanning direction, and the corresponding compensation value is determined to be -5 based on the difference in propagation distance, then the adjusted echo amplitude is 75; combined with the corresponding increasing change direction, the echo compensation parameter for the position to be repaired corresponding to this reference echo position is determined to be 75. The echo compensation parameters for each position to be repaired within the defect area are calculated using the above method.

[0038] Methods for generating target reconstructed images include: The distribution information of the reflective interface in the initial restoration image is obtained, and the corresponding target extension direction is determined according to the extension direction of the reflective interface in the non-defective region. The direction of the reflective interface in the restoration region is matched with the target extension direction. When there is a directional offset of the reflective interface in the restoration region, the echo change relationship at the corresponding position is adjusted according to the target extension direction. The reflective interface in the restoration region is made to be continuous with the reflective interface in the non-defective region, and the target reconstruction image is generated.

[0039] In this embodiment, when performing reflective interface correction on the initial repair image, the echo change information corresponding to different locations in the initial repair image is first obtained. The distribution location of the underground medium reflective interface is determined based on the echo amplitude change relationship between adjacent locations. Specifically, locations with significant and continuous echo amplitude changes are taken as reflective interface locations, and the distribution information of reflective interfaces in the initial repair image is determined based on the connection relationship between multiple continuous reflective interface locations. For example, in the initial repair image, if the echo amplitude of continuous locations within the horizontal coordinate range of 20 to 50 increases from 40 to 90, and this change area is continuously distributed along an oblique direction, then this area is determined as the corresponding underground medium reflective interface, and its corresponding location distribution is recorded.

[0040] Furthermore, the target extension direction is determined based on the extension direction of the reflective interface in the non-defective region. Specifically, the positions of continuous reflective interfaces in the non-defective region surrounding the repair area are obtained, and the extension direction of the reflective interface is determined based on the coordinate change relationship between adjacent reflective interface positions. This direction is then used as the target extension direction corresponding to the repair area.

[0041] After determining the target extension direction, the direction of the reflective interface in the repair area is matched with the target extension direction. Specifically, the position of the reconstructed reflective interface in the repair area is obtained, and the current reflective interface direction is determined based on the coordinate change relationship between adjacent reflective interface positions in the repair area. This direction is then compared with the target extension direction determined for the non-defective area. For example, if the target extension direction corresponds to a coordinate change relationship of 10 units horizontally and 5 units deep, while the reflective interface in the repair area corresponds to a change relationship of 10 units horizontally and 8 units deep, then it is determined that the reflective interface in the repair area has an directional offset.

[0042] When a deviation in the direction of the reflective interface in the repair area is detected, the echo variation relationship at the corresponding position is adjusted according to the target extension direction. Specifically, the echo amplitude variation relationship between adjacent positions in the repair area is redefined according to the target extension direction. The echo amplitude corresponding to the deviation position is adjusted so that the positional change trend of the reflective interface in the repair area corresponds to that of the reflective interface in the non-defective area, generating a target reconstruction image. For example, if the original positions of a continuous reflective interface in the repair area are (60,40), (70,48), and (80,56), while the target extension direction corresponding to the non-defective area is (60,40), (70,45), and (80,50), then the echo variation relationship at the corresponding position in the repair area is adjusted according to the target extension direction so that the reflective interface is adjusted to a continuous extension state with the surrounding area, generating a target reconstruction image.

[0043] Methods for obtaining echo continuity in a target reconstructed image include: The target reconstructed image is divided into several detection regions according to a preset spatial range, and the echo change sequence corresponding to each detection region is obtained; the echo continuity relationship between different detection regions is determined based on the echo change difference between adjacent detection regions; the echo propagation direction and continuous change region in the target reconstructed image are determined based on the echo continuity relationship.

[0044] In this embodiment, when performing echo continuity analysis on the reconstructed target image, the image is first divided into regions according to a preset spatial range. Specifically, based on the lateral scanning direction and depth direction of the radar profile image, the reconstructed target image is divided into multiple detection regions of fixed size, and a corresponding positional relationship is established for each detection region. Subsequently, the echo amplitude values ​​corresponding to different positions within each detection region are read sequentially and arranged according to the radar scanning direction to form an echo change sequence corresponding to each detection region.

[0045] Furthermore, based on the differences in echo changes between adjacent detection areas, the continuity relationship of echoes between different detection areas is determined. Specifically, the echo change sequences corresponding to adjacent detection areas are compared to obtain the direction, range, and trend of echo amplitude changes between adjacent areas. When the echo changes between adjacent areas maintain the same pattern, a continuity relationship is determined between the two detection areas.

[0046] After determining the echo continuity relationship between each detection region, the echo propagation direction and continuously changing regions in the reconstructed image are determined based on the positional connection relationship between the continuous detection regions. Specifically, continuously associated detection regions are connected according to their spatial location, the echo propagation direction is determined based on the connection direction, and the continuously connected detection regions are combined to form continuously changing regions. For example, if detection regions A, B, and C are arranged sequentially along the horizontal direction, and the echo continuity relationship is satisfied between A and B, and between B and C, then the echo propagation direction is determined to be an extension from A to C based on the region arrangement direction, and the corresponding ranges of detection regions A, B, and C are determined as continuously changing regions.

[0047] Methods for determining the correlation of reflection structures based on echo continuity include: The echo continuity relationship in the target reconstructed image is invoked to trace the echo propagation path in the continuously changing region; the extension direction, connection length and spatial adjacency relationship of the corresponding reflection structure are obtained according to the echo propagation direction; and the correlation relationship of the reflection structure in the target reconstructed image is established based on the extension direction, connection length and spatial adjacency relationship.

[0048] In this embodiment, when establishing the correlation relationship of the reflection structure, the echo continuity relationship in the target reconstructed image is first invoked to obtain the determined continuously changing regions. Then, the echo propagation path within the continuously changing regions is traced according to the connection order corresponding to the echo continuity relationship. Specifically, based on the echo connection relationship between adjacent detection regions, the positions through which the echo propagation passes are determined sequentially, and the continuously changing echo positions are connected to form the corresponding echo propagation path. For example, if there is a continuous echo relationship between detection regions A, B, and C in the target reconstructed image, and the echo change direction is from A to C, then the continuous echo positions in each detection region are connected in the order A→B→C to obtain the corresponding echo propagation path.

[0049] The extension direction, connection length, and spatial adjacency of the corresponding reflecting structure are obtained based on the echo propagation direction. Specifically, the extension direction of the reflecting structure is determined by the coordinate changes of adjacent positions along the echo propagation path; the connection length is obtained by accumulating the distances between consecutive echo positions along the echo propagation path; and the spatial adjacency is determined by the positional distances and connection states between different reflecting structures.

[0050] After obtaining the extension direction, connection length, and spatial adjacency of the corresponding reflection structures, the above information of each reflection structure is associated and recorded to establish the reflection structure association relationship in the target reconstructed image. The reflection structure association relationship includes the positional relationship, extension direction relationship, length relationship, and adjacent connection relationship of each reflection structure. For example, if there are reflection structures R1 and R2 in the target reconstructed image, R1 extends in the lower right direction with a connection length of 30 pixels, and R2 extends in the same direction as R1, and the starting position of R2 is adjacent to the ending position of R1, then a continuous association relationship is established between R1 and R2, forming the corresponding reflection structure association relationship.

[0051] Methods for verifying the consistency of reflection structure correlations include: The reflection structure association relationship is invoked to obtain the connection state of the reflection structure corresponding to the repaired region; the extension direction, connection length and spatial adjacency relationship of the reflection structure in the repaired region are compared with the corresponding reflection structure in the non-defected region; when the connection state of the reflection structure in the repaired region maintains the correspondence with the reflection structure in the non-defected region, it is determined that the target reconstructed image meets the structural association verification requirements.

[0052] In this embodiment, when performing structural association verification on the target reconstructed image, the established reflection structure association relationship is first invoked to obtain the connection status corresponding to each reflection structure within the repair area. Specifically, based on the structure number, positional relationship, and connection information in the reflection structure association relationship, it is determined whether there is a continuous connection, adjacent connection, or directional continuity relationship between each reflection structure within the repair area, and the corresponding reflection structure connection status is formed.

[0053] Furthermore, the extension direction, connection length, and spatial adjacency of the reflective structures in the repaired region are compared with the corresponding reflective structures in the non-damaged region. Specifically, based on the positional correspondence between the repaired and non-damaged regions, reference reflective structures adjacent to the repaired region in the non-damaged region are determined, and their corresponding extension directions, connection lengths, and spatial adjacency are obtained. These structural information items are then matched and judged. For example, if reflective structure A in the repaired region extends along the lower right direction with a connection length of 30 pixels, and its corresponding non-damaged region reflective structure B also extends along the lower right direction with a connection length of 32 pixels, and both maintain connections with adjacent reflective structures, then a correspondence is determined between reflective structure A and reflective structure B.

[0054] After comparing all structural information, the connection status between the reflective structures is used to determine whether the reflective structures in the repaired region and the non-missing region maintain a corresponding relationship. If the extension direction, connection length, and spatial adjacency of the reflective structures in the repaired region match the corresponding reflective structures in the non-missing region, then the target reconstructed image is determined to meet the structural association verification requirements. For example, if the connection relationship of the reflective structures in the repaired region is R1 connecting R2 and R2 connecting R3, and the corresponding connection relationship of the reflective structures in the non-missing region is also B1 connecting B2 and B2 connecting B3, and the extension direction and connection length of each structure are within the corresponding range, then the target reconstructed image is determined to meet the structural association verification requirements.

[0055] Methods for outputting the final repair results include: Based on the reflection structure correlation verification results, the echo change continuous parameters and structure extension correlation parameters corresponding to the target reconstructed image are obtained; the echo change continuous parameters and structure extension correlation parameters are combined and calculated to obtain the reconstruction consistency evaluation results, and the reconstruction consistency evaluation results are compared with the preset consistency conditions; when the reconstruction consistency evaluation results meet the preset consistency conditions, the target reconstructed image is determined as the final repair result and output.

[0056] Based on the echo continuity relationship between detection regions in the reconstructed image, the direction and trend of echo amplitude changes between adjacent detection regions are statistically analyzed to generate echo change continuity parameters. Simultaneously, based on the reflection structure association verification results, the extension direction, connection length, and spatial adjacency relationship of the corresponding reflection structures in the repaired and non-defective regions are obtained to generate structure extension association parameters. For example, if the echo amplitude sequences of adjacent detection regions in the reconstructed image are {50, 55, 60} and {65, 70, 75}, and the echoes in both regions change in an increasing direction, then the corresponding echo change continuity parameters are determined based on this relationship. Furthermore, if the reflection structure in the repaired region has a length of 30 pixels, extends in the lower right direction, and remains connected to the corresponding reflection structure in the non-defective region, then the structure extension association parameters are determined based on the aforementioned structural information.

[0057] Furthermore, the echo variation continuity parameter and the structural extension correlation parameter are combined for calculation to obtain the reconstruction consistency evaluation result. Specifically, the parameters representing the echo continuity relationship and the parameters representing the structural correlation relationship are calculated according to a preset combination method to generate an evaluation result representing the consistency of the target reconstructed image.

[0058] For example, obtaining the echo variation continuity parameter C1=0.90 and the structural extension correlation parameter C2=0.85, and according to the preset calculation method: C=C1×C2, the reconstruction consistency evaluation result C=0.765 is obtained. After obtaining the reconstruction consistency evaluation result, this result is compared with the preset consistency condition. When the reconstruction consistency evaluation result meets the preset consistency condition, the corresponding target reconstruction image is determined as the final repair result and output. For example, if the preset consistency condition is set to the reconstruction consistency evaluation result being no less than 0.70, and the calculated evaluation result is 0.765, the target reconstruction image is determined to meet the consistency condition, and this target reconstruction image is output as the final repair result. Example

[0059] Please see Figure 2 As shown, for parts not described in detail in this embodiment, please refer to the description in Embodiment 1. A ground-penetrating radar image defect repair and reconstruction processing system is provided, comprising: The defect structure perception module is used to acquire radar profile images generated by ground penetrating radar, locate defect areas in the images, and generate defect structure description information based on the spatial adjacency relationship between the boundary of the defect area and the surrounding effective echo area, and the echo change trend. The echo trajectory association module is used to call the description information of the defect structure, perform association analysis on the radar echo propagation trajectories of corresponding positions on both sides of the defect area, and determine the echo trajectory compensation path passing through the defect area based on the continuous propagation relationship of the echoes. The defect area reconstruction module is used to spatially reconstruct the echo information within the defect area based on the echo trajectory compensation path and the correspondence between the defect location and the associated echo trajectory, thereby generating an initial repair image. The medium boundary correction module is used to call the initial repair image and, based on the continuous extension relationship of the underground medium reflection boundary, perform consistency correction on the reflection interface between the repair area and the non-defective area to generate the target reconstruction image; The multi-scale echo verification module is used to acquire the echo continuity relationship in the target reconstructed image, determine the correlation relationship of the reflection structure based on the echo continuity relationship, and perform consistency verification on the correlation relationship of the reflection structure. When the preset consistency conditions are met, the final repair result is output. The various modules are connected via wired and / or wireless means to realize data transmission between modules. Example

[0060] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the intelligent repair and reconstruction processing method for ground penetrating radar defect images provided above.

[0061] Since the electronic device described in this embodiment is the electronic device used to implement the intelligent repair and reconstruction processing method for damaged images of ground penetrating radar according to the embodiments of this application, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the intelligent repair and reconstruction processing method for damaged images of ground penetrating radar described in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. Any electronic device used by those skilled in the art to implement the intelligent repair and reconstruction processing method for damaged images of ground penetrating radar according to the embodiments of this application falls within the scope of protection of this application.

[0062] It should be noted that all formulas in this manual are calculated by removing dimensions and taking their numerical values. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0063] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for intelligent repair and reconstruction of damaged images from ground-penetrating radar, characterized in that, include: S1. Acquire radar profile images generated by ground-penetrating radar detection, locate the defective areas in the images, and generate defective structure description information based on the spatial adjacency relationship between the boundary of the defective area and the surrounding effective echo area, and the echo change trend. S2. Call the description information of the defective structure, perform correlation analysis on the radar echo propagation trajectory at the corresponding positions on both sides of the defective area, and determine the compensation path of the echo trajectory passing through the defective area based on the continuous propagation relationship of the echoes. S3. Based on the echo trajectory compensation path, according to the correspondence between the defect location and the associated echo trajectory, the echo information in the defect area is spatially reconstructed to generate an initial repair image. S4. Call the initial repair image, and based on the continuous extension relationship of the underground medium reflection boundary, perform consistency correction on the reflection interface between the repair area and the non-defective area to generate the target reconstruction image; S5. Obtain the echo continuity relationship in the target reconstructed image, determine the reflection structure association relationship based on the echo continuity relationship, and verify the consistency of the reflection structure association relationship. When the preset consistency condition is met, output the final repair result.

2. The intelligent repair and reconstruction method for damaged images of ground-penetrating radar according to claim 1, characterized in that, The method for generating description information of the defective structure includes: The radar profile image is divided into regions to obtain the set of missing regions corresponding to continuous missing regions, and the set of boundary points on the boundary of each missing region is extracted; based on the spatial distance relationship between the set of boundary points and the adjacent effective echo regions, the associated regions corresponding to the missing regions are determined. The system acquires the direction of echo amplitude variation, the trend of continuous echo variation, and the direction of boundary extension within the associated region. It establishes the spatial relationship between the defect area and the surrounding effective echo area, and generates defect structure description information including the defect range, boundary location, and echo extension direction.

3. The method for intelligent repair and reconstruction of damaged images from ground-penetrating radar according to claim 2, characterized in that, The method for determining the echo trajectory compensation path based on the defect structure description information includes: The system retrieves the defect boundary location and echo propagation direction from the defect structure description information to match the effective echo trajectories at corresponding locations on both sides of the defect area; it obtains the echo change relationship between adjacent scanning positions according to the radar scanning direction to determine the degree of continuity between different effective echo trajectories; it filters candidate echo trajectories that meet the continuous echo propagation condition based on the degree of continuity, and determines the candidate echo trajectories as the echo trajectory compensation path that passes through the defect area.

4. The intelligent repair and reconstruction method for damaged images of ground-penetrating radar according to claim 3, characterized in that, The method for generating the initial repaired image includes: The reference echo position corresponding to each location to be repaired within the defect area is determined based on the echo trajectory compensation path; the echo amplitude, direction of change, and propagation distance information corresponding to the reference echo position are obtained; the echo compensation parameters are determined based on the correlation between the reference echo position and the location to be repaired; the echo information of each location to be repaired within the defect area is reconstructed based on the echo compensation parameters, so that the trend of the reconstructed echo change is consistent with the echo trajectory compensation path, and an initial repair image is generated.

5. The intelligent repair and reconstruction method for damaged images of ground-penetrating radar according to claim 4, characterized in that, The method for determining the echo compensation parameters includes: Obtain the propagation distance, echo change direction, and spatial relationship corresponding to each reference echo position; determine the corresponding distance compensation parameters based on the propagation distance differences between different reference echo positions and the position to be repaired; adjust the echo amplitude corresponding to each reference echo position based on the distance compensation parameters and echo change direction to generate the echo compensation parameters corresponding to the position to be repaired.

6. The intelligent repair and reconstruction method for damaged images of ground-penetrating radar according to claim 5, characterized in that, The method for generating the target reconstructed image includes: The distribution information of the reflective interface in the initial restoration image is obtained, and the corresponding target extension direction is determined according to the extension direction of the reflective interface in the non-defective region. The direction of the reflective interface in the restoration region is matched with the target extension direction. When there is a directional offset of the reflective interface in the restoration region, the echo change relationship at the corresponding position is adjusted according to the target extension direction. The reflective interface in the restoration region is made to be continuous with the reflective interface in the non-defective region, and the target reconstruction image is generated.

7. The intelligent repair and reconstruction method for damaged images of ground-penetrating radar according to claim 6, characterized in that, The method for obtaining the echo continuity relationship in the target reconstructed image includes: The target reconstructed image is divided into several detection regions according to a preset spatial range, and the echo change sequence corresponding to each detection region is obtained; the echo continuity relationship between different detection regions is determined based on the echo change difference between adjacent detection regions; the echo propagation direction and continuous change region in the target reconstructed image are determined based on the echo continuity relationship.

8. The intelligent repair and reconstruction method for damaged images of ground-penetrating radar according to claim 7, characterized in that, The method for determining the correlation of reflection structures based on echo continuity includes: The echo continuity relationship in the target reconstructed image is invoked to trace the echo propagation path in the continuously changing region; the extension direction, connection length and spatial adjacency relationship of the corresponding reflection structure are obtained according to the echo propagation direction; and the correlation relationship of the reflection structure in the target reconstructed image is established based on the extension direction, connection length and spatial adjacency relationship.

9. The intelligent repair and reconstruction method for damaged images of ground-penetrating radar according to claim 8, characterized in that, The method for verifying the consistency of the correlation between reflection structures includes: The reflection structure association relationship is invoked to obtain the connection state of the reflection structure corresponding to the repaired region; the extension direction, connection length and spatial adjacency relationship of the reflection structure in the repaired region are compared with the corresponding reflection structure in the non-defected region; when the connection state of the reflection structure in the repaired region maintains the correspondence with the reflection structure in the non-defected region, it is determined that the target reconstructed image meets the structural association verification requirements.

10. The intelligent repair and reconstruction method for damaged images of ground-penetrating radar according to claim 9, characterized in that, The method for outputting the final repair result includes: Based on the reflection structure correlation verification results, the echo change continuous parameters and structure extension correlation parameters corresponding to the target reconstructed image are obtained; the echo change continuous parameters and structure extension correlation parameters are combined and calculated to obtain the reconstruction consistency evaluation results, and the reconstruction consistency evaluation results are compared with the preset consistency conditions; when the reconstruction consistency evaluation results meet the preset consistency conditions, the target reconstructed image is determined as the final repair result and output.

Images