Method and device for removing redundant data of ultra-wideband penetration imaging radar

Abstract

The invention discloses a method and device for removing redundant data of an ultra-wideband penetrating imaging radar, and belongs to the technical field of ultra-wideband radar signal processing. The method comprises the following steps: collecting original mobile detection data, and carrying out differential processing in an advancing direction; carrying out image edge detection on the movement detection data subjected to differential processing along the advancing direction to obtain an edge detection result of the movement detection signal; and carrying out accumulative integration on the edge detection result along the depth direction, generating a depth-direction energy distribution curve, setting a threshold value based on the energy distribution curve, removing static redundant data of the ultra-wideband radar, and retaining effective mobile detection data. According to the method, the image recognition technology is adopted, and the median filtering is combined, so that the working efficiency and the recognition precision of removing the repeated redundant data are greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of ultra-wideband radar signal processing, and in particular relates to a method and a device for removing redundant data of an ultra-wideband penetrating imaging radar. Background Art

[0002] Ultra-wideband penetrating imaging radar is widely used in many fields such as urban road disease detection, tunnel advance prediction, dam disease detection, engineering exploration and site selection, and lunar and deep space exploration due to its advantages of non-destructiveness, convenience and high efficiency. Ultra-wideband penetrating imaging radar transmits ultra-wideband electromagnetic waves in space. When encountering an abnormal body, that is, an interface with discontinuous dielectric constant, reflection and scattering will occur. After the radar receiving antenna receives the reflected and scattered signals, the corresponding detection data is obtained after amplification and sampling by the receiver. By analyzing, processing and imaging the detection data, underground structure information such as road cavities, broken zones, water leakage areas, geological stratification, etc. on the detection route is obtained. Ultra-wideband penetrating imaging radar generally works in a mobile manner to obtain continuous profiles of the detection area. There are usually two ways to transmit signals: timed trigger and fixed distance trigger. Ground-coupled radars that are in direct contact with the ground can use measuring wheels and other means to obtain equally spaced detection signals by fixed distance triggering, with basically no redundant data. However, in some special scenarios, direct contact is not possible, or it is difficult to directly contact the detection interface, so the only way to detect is by air coupling. At this time, the use of timed triggering to transmit signals will inevitably generate repeated redundant signals when the device is in a static state. These repeated redundant signals must be removed in the later data analysis, otherwise it will bring a lot of workload, especially in the case of massive data. The existing methods for removing repeated redundant data from ultra-wideband penetrating imaging radars are usually manual removal or cross-correlation methods. The manual removal method has a huge workload and is not accurate enough, which may cause under-removal and over-removal. Although the cross-correlation method improves efficiency, it also has insufficient accuracy, especially when the radar moves slowly and the similarity of adjacent data is high. Summary of the invention

[0003] To solve the above technical problems, the present invention provides a method and device for removing redundant data of ultra-wideband penetrating imaging radar, which adopts image recognition technology and combines median filtering to greatly improve the work efficiency and recognition accuracy of removing repeated redundant data.

[0004] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0005] A method for removing redundant data of an ultra-wideband penetrating imaging radar, the method comprising:

[0006] Step 1: Collect the original mobile detection data and perform differential processing in the direction of travel;

[0007] Step 2: Perform image edge detection on the differential processed motion detection data along the moving direction to obtain an edge detection result of the motion detection signal;

[0008] Step 3: Accumulate and integrate the edge detection results along the depth direction to generate a depth energy distribution curve, and set a threshold based on the energy distribution curve to remove the static redundant data of the ultra-wideband radar and retain the effective mobile detection data.

[0009] In another aspect, the present invention provides a device for removing redundant data of ultra-wideband penetrating imaging radar, comprising:

[0010] The acquisition module is used to collect the original mobile detection data and perform differential processing in the direction of travel;

[0011] A detection module is used to perform image edge detection along the traveling direction on the motion detection data after differential processing to obtain an edge detection result of the motion detection signal;

[0012] The removal module is used to accumulate and integrate the edge detection results along the depth direction to generate a depth energy distribution curve, and set a threshold based on the energy distribution curve to remove the static redundant data of the ultra-wideband radar and retain the effective mobile detection data.

[0013] In a third aspect, the present invention provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned method for removing redundant data of ultra-wideband penetrating imaging radar.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the aforementioned method for removing redundant data of ultra-wideband penetrating imaging radar.

[0015] The beneficial effects of the present invention are:

[0016] The present invention improves the processing accuracy through 0 / 1 binary quantization of the difference components between adjacent frames and depth-wise energy integration, and uses the median filtering method to remove unexpected spikes and random noise, thereby improving accuracy; all processing can be automatically handled by the program, taking into account both accuracy and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flow chart of a method for removing redundant data of ultra-wideband penetrating imaging radar according to the present invention;

[0018] Figure 2is the original motion detection data curve;

[0019] Figure 3 is the forward differential result curve;

[0020] Figure 4 Schematic diagram for Sobel image edge detection window selection;

[0021] Figure 5 Schematic diagram of operators for image edge detection in horizontal and vertical directions;

[0022] Figure 6 This is a schematic diagram of the edge detection result of the Sobel image in the moving direction;

[0023] Figure 7 It is a schematic diagram of the energy integration result in depth;

[0024] Figure 8 The integrated energy curve of the Sobel detection result in depth;

[0025] Fig. 9 Schematic diagram for threshold selection;

[0026] Fig.10 It is the effective moving data curve after removing redundant static data. DETAILED DESCRIPTION

[0027] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0028] like Figure 1 As shown in FIG. 1 , a flow chart of a method for removing redundant data of ultra-wideband penetrating imaging radar according to the present invention is shown, which specifically includes the following steps:

[0029] Step 1: Collect the original motion detection data and perform differential processing in the moving direction. Subtract the data of adjacent frames to remove the similar components of adjacent frames and retain the difference components.

[0030] When the ultra-wideband penetrating imaging radar detects in situ, the detection signals are highly similar because the surrounding detection targets have not changed, such as Figure 2 As shown in the figure, the black dotted box represents the motion detection signal, and the outside of the dotted box is the static / redundant signal. It can be seen from the figure that the static / redundant signal has a very high similarity between adjacent frames. By differentiating adjacent frames, the similar components can be removed, highlighting the motion detection signal; that is, by differentiating in the direction of travel, that is, by subtracting adjacent frame data, the similar components of adjacent frames can be removed and the difference components can be retained. The processing results are shown in Figure 3 As shown, the black dotted box in the figure represents the motion detection data, and it can be seen that the difference components between adjacent frames have been highlighted.

[0031] Step 2: Perform image edge detection on the differentiated motion detection data, perform 0 / 1 binary quantization on the difference component, select a depth window to highlight the difference in the image, perform image edge detection along the moving direction, and obtain the edge detection result of the motion detection signal;

[0032] The Sobel operator is used to detect image edges on the differentiated detection data, that is, the difference component is quantized into 0 / 1 binary values. The specific method is as follows: In order to improve the accuracy of edge detection, the area with the most prominent depth difference is first selected as the window for Sobel image edge detection, such as Figure 4 As shown in the black dotted box marked with 2, the differences in the image are further highlighted. Specifically, there are 3 dotted boxes in the figure: Window No. 1 cannot be selected because the original signal intensity is large and the residual basis is large after the forward differentiation, which fails to highlight the difference between the static data and the moving data; Window No. 3 cannot be selected because the original signal signal-to-noise ratio is low and the difference between the static data and the moving data is also not highlighted after the forward differentiation; Window No. 2 has a high original signal signal-to-noise ratio. After the forward differentiation, the difference between the static data and the moving data is very obvious, which makes it easy to detect the image edge, so Window No. 2 is selected; Then, Sobel image edge detection is performed along the forward direction. The Sobel operator detects edges based on the phenomenon that the weighted difference of the grayscale of the upper and lower and left and right neighboring points of the pixel reaches an extreme value at the edge. The Sobel operator contains two sets of 3×3 filters, which are sensitive to edges in the horizontal and vertical directions respectively. The Sobel operators in the horizontal and vertical directions are shown as follows Figure 5 As shown (horizontally on the left and vertically on the right), the two operators are then used to perform convolution operations on the input image to obtain the convolution results of each point in the horizontal and vertical directions, i.e., the horizontal gradient and vertical gradient , and then do a square sum operation: , select the point whose gradient G is greater than a certain threshold as the edge point of (x, y). Figure 2 It can be seen from the dotted window 2 that the difference in the image moving direction (horizontal direction) is more obvious, so the moving direction is selected for edge detection. The detection results are as follows: Figure 6 As shown, from Figure 6 As can be seen in the figure, the signal detected by the radar in motion appears bright white, which is significantly different from the static / redundant data.

[0033] Preferably, when quantizing the difference components between adjacent frames, in addition to the Sobel operator for image edge detection, the gradient operator, Roberts operator, Prewitt operator, Laplacian operator, etc. can also be used as a replacement. There are certain differences in the quantization results of different operators, but it does not affect the final effect of removing redundant data.

[0034] Step 3: Accumulate and integrate the image edge detection results along the depth direction to generate a depth energy distribution curve, and set a reasonable threshold based on the energy distribution curve to remove the static redundant data of the ultra-wideband radar and retain the effective mobile detection data;

[0035] The result of the Sobel image edge detection in the traveling direction is accumulated and integrated along the depth direction. The result is as follows: Figure 7 As shown in the figure, the numbers 1-5 marked in the figure represent multiple "thin" spikes and glitches, which will affect the accuracy of extracting motion detection data. The energy integration formula is as follows, where E is the energy after integration of each frame of data, to The window selected for depth edge detection:

[0036] .

[0037] When the ultra-wideband penetrating imaging radar is used for field detection, unexpected spikes and random noise may appear in the detection data profile due to many reasons, such as improper placement of the antenna, battery replacement, restart, etc., or Figure 7 The spikes and glitches (salt and pepper noise) shown in the figure can be removed by median filtering (i.e., mutation points). Median filtering is a nonlinear image processing technique that is widely used to remove noise from images, and is particularly suitable for removing salt and pepper noise. The working principle of median filtering is to move through each pixel of the image through a window of adjustable size. For each pixel in the window, the pixel values ​​are sorted, and the middle value of the sorting is used to replace the value of the center pixel of the original window. This method is particularly effective because it is not affected by extreme values, which is especially important when processing images containing salt and pepper noise. Compared with linear filters, median filtering performs better in maintaining edge information. A sliding window of 20 points is used to filter the energy curve after depth integration, and the results are shown in the figure. Figure 8 As shown in the figure, the 1~5 bands marked in the figure are the data that are finally selected to be retained. Figure 2 The motion detection signals in the black dashed box are highly consistent.

[0038] Based on the energy after median filtering, set a reasonable threshold, such as Fig. 9 As shown in the figure, the dotted line is the selected threshold. Data exceeding the threshold is a mobile detection signal and is retained; data below the threshold is interference, which will affect the accuracy of extracting mobile detection data. The static redundant data (i.e., interference) of the ultra-wideband radar is removed and the valid mobile detection data is retained. The results are shown in Fig.10 shown.

[0039] On the other hand, the present invention provides a device for removing redundant data of ultra-wideband penetrating imaging radar, and each module included in the device can implement each step of the aforementioned method, specifically including:

[0040] The acquisition module is used to collect the original mobile detection data and perform differential processing in the direction of travel;

[0041] A detection module is used to perform image edge detection along the traveling direction on the motion detection data after differential processing to obtain an edge detection result of the motion detection signal;

[0042] The removal module is used to accumulate and integrate the edge detection results along the depth direction to generate a depth energy distribution curve, and set a threshold based on the energy distribution curve to remove the static redundant data of the ultra-wideband radar and retain the effective mobile detection data.

[0043] In a third aspect, the present invention provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned method for removing redundant data of ultra-wideband penetrating imaging radar.

[0044] In a fourth aspect, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the aforementioned method for removing redundant data of ultra-wideband penetrating imaging radar.

[0045] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for removing redundant data of ultra-wideband penetrating imaging radar, characterized in that: The method comprises: Step 1: Collect the original mobile detection data and perform differential processing in the direction of travel; Step 2: Perform image edge detection on the differential processed motion detection data along the moving direction to obtain an edge detection result of the motion detection signal; Step 3: Accumulate and integrate the edge detection results along the depth direction to generate a depth energy distribution curve, and set a threshold based on the energy distribution curve to remove the static redundant data of the ultra-wideband radar and retain the effective mobile detection data.

2. The method for removing redundant data of ultra-wideband penetrating imaging radar according to claim 1, characterized in that: In the step 1, performing differential processing in the moving direction includes removing similar components of adjacent frames by subtracting data of adjacent frames and retaining difference components.

3. The method for removing redundant data of ultra-wideband penetrating imaging radar according to claim 1, characterized in that: In the step 2, the Sobel operator is used to perform image edge detection on the differential processed motion detection data, and the difference component is quantized to 0 / 1 binary value, which specifically includes: selecting a depth window, performing Sobel image edge detection along the moving direction, and obtaining the edge detection result of the motion detection signal.

4. The method for removing redundant data of ultra-wideband penetrating imaging radar according to claim 3, characterized in that: The Sobel operator can be replaced by any one of a gradient operator, a Roberts operator, a Prewitt operator, and a Laplacian operator.

5. The method for removing redundant data of ultra-wideband penetrating imaging radar according to claim 1, characterized in that: In step 3, a median filtering method is used to remove the static redundant data of the ultra-wideband radar.

6. The method for removing redundant data of ultra-wideband penetrating imaging radar according to claim 5, characterized in that: The median filtering method uses a 20-point sliding window to filter the depth energy distribution curve.

7. The method for removing redundant data of ultra-wideband penetrating imaging radar according to claim 6, characterized in that: The static redundant data of the ultra-wideband radar includes unexpected peaks and random noises existing in the detection data profile.

8. A device for removing redundant data of ultra-wideband penetrating imaging radar, characterized in that: include: The acquisition module is used to collect the original mobile detection data and perform differential processing in the direction of travel; A detection module is used to perform image edge detection along the traveling direction on the motion detection data after differential processing to obtain an edge detection result of the motion detection signal; The removal module is used to accumulate and integrate the edge detection results along the depth direction to generate a depth energy distribution curve, and set a threshold based on the energy distribution curve to remove the static redundant data of the ultra-wideband radar and retain the effective mobile detection data.

9. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; Wherein, when one or more programs are executed by the one or more processors, the one or more processors implement the method for removing redundant data of ultra-wideband penetrating imaging radar as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: Executable instructions are stored thereon, and when the instructions are executed by the processor, the processor can implement the method for removing redundant data of ultra-wideband penetrating imaging radar as described in any one of claims 1 to 7.

Images