Microwave imaging three-dimensional image processing method

Abstract

The invention discloses a microwave imaging three-dimensional image processing method, and belongs to the technical field of detection, identification and positioning application of pipelines in walls. The method comprises the following steps: obtaining a scattering echo signal from a target, obtaining a three-dimensional amplitude image, and carrying out amplitude compensation on the distance dimension of the three-dimensional amplitude image; spatial convolution calculation is carried out on the three dimensions to obtain a filtered image; performing gamma transformation and logarithmic transformation on the filtered image pixel by pixel; performing projection conversion on the distance dimension of the logarithmic transformation image to obtain a projection image and a distance dimension maximum projection coordinate; obtaining a color gradation index value corresponding to each pixel point of the projection image to obtain a pseudo-color image; filling the three-dimensional image with the position of the projection image at the projection coordinate of the distance dimension maximum value; and the pseudo-color image is sent to a pipeline detection, identification and positioning software module for identification and positioning. According to the method, the three-dimensional outline of the pipeline can be highlighted, and subsequent pipeline detection, recognition and positioning are facilitated.

Description

Technical Field

[0001] The invention belongs to the technical field of wall internal pipeline detection, identification and positioning application, and relates to a microwave imaging three-dimensional image processing method. Background Art

[0002] With the increasing development of the economy and society, new commercial housing and commercial centers have sprung up like mushrooms after a spring rain, while old buildings and commercial buildings in the city center are also facing upgrading and renovation projects, such as installing elevators, redecorating, and re-laying water and electricity pipelines. The actual problem is that the original construction drawings have not been properly preserved or partially lost due to their age. At this time, when the water and electricity pipelines are upgraded and renovated, it is difficult to locate the original pipelines. Therefore, a technical means is needed that can penetrate the wall and accurately locate and identify the pipeline position.

[0003] Microwave, millimeter wave, and terahertz three-dimensional imaging technologies can be applied to technical fields such as internal defect detection of composite materials, missing items detection in packaging boxes, geological mineral exploration, and detection and positioning of pipelines inside walls. The electromagnetic waves emitted by microwave imaging systems operating in the 3~12GHz frequency band can easily penetrate walls and image the inside of walls or penetrate walls to image the inside of rooms, and the resulting images are three-dimensional images. In the application scenario of pipeline detection and positioning inside walls, the equipment users are concerned about the spatial position of the pipelines inside the wall, and the resulting three-dimensional images contain the signals of the wall itself and the pipeline signals mixed together. How to effectively filter out the clutter interference caused by the wall, highlight the effective outline of the pipeline, and provide high-quality images for subsequent pipeline identification and positioning has become a key technology worth solving. Summary of the invention

[0004] The technical problem to be solved by the present invention is: a microwave imaging three-dimensional image processing method is provided for the application field of pipeline detection, identification and positioning inside the wall. The present invention proposes a method for three-dimensional processing of three-dimensional images formed by a microwave imaging system in the technical field of pipeline detection, identification and positioning inside the wall, which can effectively filter out the clutter interference of the wall, improve the contrast of the pipeline outline, and display the pipeline outline in three dimensions.

[0005] The present invention provides a microwave imaging three-dimensional image processing method, comprising the following steps: S110, after the microwave 3D imaging system obtains the scattered echo signal from the target, it obtains the slice image of each distance through the 3D image reconstruction algorithm to form a 3D complex image , modulus the 3D complex image to obtain the 3D amplitude image ; S120, for three-dimensional amplitude image The distance dimension Amplitude compensation is performed, and the image data after compensation is ; S130, using a three-dimensional spatial filter to process the distance-compensated image data of , , Dimensional spatial convolution calculation is performed to obtain the filtered image: ; S140, for the filtered image Perform pixel-by-pixel gamma transformation, and the image after gamma transformation is ; S150, for gamma transformed images Perform logarithmic transformation and obtain the logarithmic transformed image: ; S160, log-transformed image The distance dimension is projected and transformed to obtain the projected image and distance dimension maximum projection coordinates ; S170, using a rainbow color scale mapping table, projects the coordinates of the maximum value of the distance dimension Index the rainbow color scale mapping table to obtain the projected image The color index value corresponding to each pixel point finally obtains the pseudo-color image ; S180, project the image The position of the maximum projection coordinate in the distance dimension Fill it into the 3D amplitude image to form a 3D point cloud image , the image will be projected Mapped to coordinates The pixel points are converted into a 3D point cloud image and then sent to the host computer software for 3D display using visualization tools. S190, pseudo color image The pipeline is sent to the pipeline detection, identification and positioning software module for identification and positioning, and the positioning information is sent to the host computer for online display.

[0006] Compared with the prior art, the present invention has the following beneficial effects: The present invention is directed to the field of microwave three-dimensional imaging technology, and is specifically applied to the application scenarios of pipeline detection, identification and positioning inside walls. When the three-dimensional image formed contains effective pipeline signals and wall clutter interference signals, an image processing method is required to suppress clutter and enhance the pipeline contour contrast; the current image processing method generally adopts a two-dimensional image processing method, and the two-dimensional image processing is difficult to effectively process the three-dimensional contour of the target, and cannot highlight the three-dimensional contour of the target. At the same time, the information of the distance dimension is easily lost, so the effective information of the processed image is lost. In the distance compensation module of the present invention, the attenuation of the echo signal caused by the difference in the plane of the target distance imaging system is effectively compensated, so that the integrity of the three-dimensional contour of the target can be highlighted, and at the same time, a three-dimensional spatial filtering method is further used to filter out isolated noise points and smooth the pixels with drastic changes in the three-dimensional directions; a gamma transformation module is used to enhance the overall contour of the pipeline, and then a logarithmic transformation method is used to filter out the wall clutter interference signal, and the pipeline contour signal with strong reflection is retained. In the pseudo-color image conversion module, the distance coordinates are associated with the color scale mapping table, so that pixels at different distances display different colors, and pipeline detection and positioning can be more easily performed in visual processing. The three-dimensional image processing technology adopted in the present invention can effectively suppress wall clutter, highlight the three-dimensional contour of the pipeline, and is more conducive to subsequent pipeline detection, identification and positioning. It is a good method worthy of promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 It is a flow chart of a microwave imaging three-dimensional image processing method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0008] The following is a detailed description of an embodiment of the present invention. This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiment.

[0009] The entire antenna array of the microwave 3D imaging system adopts a sparse array configuration design, including sparse transmitting array elements and sparse receiving array elements. The frequency band of the signal transmitted by the transceiver front end ranges from 3 to 12 GHz. Under the triggering of the host computer control timing generator, the microwave 3D imaging system turns on the transmitting array elements in time-sharing, and the receiving array elements turn on at the same time and receive the backscattered echo signals of the target.

[0010] Reference below Figure 1 The microwave imaging three-dimensional image processing method according to an embodiment of the present invention is described. Figure 1 As shown, the method comprises the following steps: S110, after the microwave three-dimensional imaging system obtains the scattered echo signal, it obtains the slice image of each distance through the three-dimensional image reconstruction algorithm to form a three-dimensional complex image , modulus the 3D complex image to obtain the 3D amplitude image , the number of pixels in the dimensions of the three-dimensional amplitude image are , , .

[0011] Among them, after the microwave three-dimensional imaging system obtains the scattered echo signal, it can perform digital down-conversion to obtain the demodulated IQ signal, and then obtain the slice image of each distance through the three-dimensional image reconstruction algorithm.

[0012] S120, for three-dimensional amplitude image The distance dimension Amplitude compensation is performed, and the image data after compensation is .

[0013] Since the target is a three-dimensional target, the target itself has a difference in distance from the position of the imaging system plane, and the echo signal of the target attenuates more when the target is farther away. Therefore, it is necessary to perform amplitude attenuation compensation in the distance dimension of the three-dimensional image.

[0014] In one embodiment, the distance dimension amplitude attenuation compensation coefficient used is ,in is a three-dimensional amplitude image The coordinates distributed in the distance dimension, is the minimum distance value divided during 3D image reconstruction, It is the maximum value of the distance divided during three-dimensional image reconstruction.

[0015] S130, using a three-dimensional spatial filter to process the distance-compensated three-dimensional image of , , Dimensional spatial convolution calculation is performed to obtain the filtered image: .

[0016] In one embodiment, the three-dimensional spatial filter The main function is to filter out isolated noise signals in the image data after distance compensation and smooth the parts with drastic pixel fluctuations. The specific design is: , in, , which are three two-dimensional spatial filters, and then combined into a three-dimensional spatial filter , , , It indicates that the three dimensions of the three-dimensional spatial filter are 3.

[0017] In one embodiment, the spatial convolution calculation includes: a three-dimensional spatial filter With 3D images Convolution is performed pixel by pixel.

[0018] In one embodiment, before performing three-dimensional spatial filtering, the distance-compensated three-dimensional image is Perform boundary zero padding operation, that is, after edge zero padding operation The data dimension becomes .

[0019] The three-dimensional spatial filtering calculation process is: , in, , , is the subscript of the pixel index of the three dimensions of the image, , , They are the indexes of the three-dimensional spatial filter in the x, y, and z dimensions respectively. , is the number of image pixels in horizontal and vertical dimensions, is the number of pixels in the distance dimension.

[0020] S140, for the filtered image Perform pixel-by-pixel gamma transformation, and the image after gamma transformation is This step can improve the integrity of the pipeline outline in the image.

[0021] S150, for gamma transformed images Perform logarithmic transformation and obtain the logarithmic transformed image: .

[0022] In one embodiment, the logarithmic transformation process may include: applying a logarithmic transformation with a base of 2 to each pixel, that is, the logarithmic transformation image is .

[0023] This step can improve the contrast of the pipeline target in the entire three-dimensional image and effectively suppress the clutter signal scattered by the wall.

[0024] S160, log-transformed image The distance dimension is projected and transformed to obtain the projected image and distance dimension maximum projection coordinates The projected image , the maximum projection coordinate of the distance dimension .in It means to find the maximum value. Indicates the distance coordinate index value corresponding to the maximum value.

[0025] S170, using a rainbow color scale mapping table, projects the coordinates of the maximum value of the distance dimension Index the rainbow color scale mapping table to obtain the projected image The color index value corresponding to each pixel point finally obtains the pseudo-color image The rainbow color scale mapping table is well known to those skilled in the art.

[0026] The color scale mapping table is expressed as ,in Represents the value of the color scale mapping table, Indicates the index value of the mapping table. Indicates the corresponding three-channel color scale value in the mapping table. Projection coordinates according to the maximum value of the distance dimension The specific values ​​of the three channels corresponding to the rainbow color scale mapping table are ,in express Coordinates in The color value indexed in express Coordinates in The color value indexed in express Coordinates in The color scale value indexed in . Therefore, the pseudo-color floating-point image of the three channels is ,in , , Respectively expressed as: , Furthermore, the three-channel floating-point images are converted into 8-bit grayscale images respectively, and the final pseudo-color normalized image is obtained as follows: ,in , , Respectively expressed as: , After the color scale mapping, different distance information in the image is represented by different color scales, which makes the image details more detailed and the target contour more prominent.

[0027] S180, project the image The position of the maximum projection coordinate in the distance dimension Fill it into the 3D amplitude image to form a 3D point cloud image , the image will be projected Mapped to coordinates Pixel points, get the three-dimensional point cloud image The generated 3D point cloud image is sent to the host computer software and displayed in 3D using a visualization tool. The visualization tool may be a VTK (Visualization Toolkit) rendering software module.

[0028] S190, pseudo color image The pipeline is sent to the pipeline detection, identification and positioning software module for detection, identification and positioning, and the positioning information is sent to the host computer for online display. In the pipeline detection, identification and positioning software module, the position in the three-dimensional space is detected, identified and positioned. The same visualization tool as step S180 can be used in the host computer for three-dimensional display.

[0029] In summary, the technical problem to be solved by the present invention is: the microwave three-dimensional imaging system needs to penetrate the wall for imaging, the system has a high resolution in the azimuth dimension and the distance dimension, and the resulting three-dimensional image also contains wall information, while the wall itself is not a part of the system and will interfere with the internal pipeline. Therefore, it is necessary to effectively process the three-dimensional image output by the imaging system, filter out the clutter interference of the wall itself, highlight the effective information of the pipeline inside the wall, and make the subsequent pipeline identification and positioning more accurate.

[0030] The process of the three-dimensional image processing method proposed in the present invention includes a distance compensation module, a spatial filtering module, a gamma transformation module, a logarithmic conversion module, a projection conversion module, a pseudo-color image conversion module and a three-dimensional point cloud rendering module. The distance of the target from the imaging system affects the amplitude of the echo signal. The distance compensation module performs amplitude compensation on the distance dimension in the three-dimensional image so that the amplitude levels of slice images at different distances remain consistent; a three-dimensional spatial filter is designed in the spatial filtering module to filter out isolated noise points in the three-dimensional image and smooth the pixels with drastic amplitude changes; the gamma transformation module enhances the overall signal of the pipeline and the wall in the image. The pipeline itself is made of metal or plastic and has good reflection characteristics for electromagnetic waves; the logarithmic transformation module enhances the contrast of the strongly reflected target information in the image and highlights the three-dimensional contour of the pipeline; the projection conversion module performs maximum value on the distance dimension of the above-mentioned processed image. Projection conversion, and obtain the coordinate position of the maximum value of the distance dimension; the pseudo-color image conversion module combines the rainbow color scale table with the two-dimensional image and distance coordinates output by the projection conversion module. Pixel points of different distance coordinates correspond to different color scale points. The outline of the pipeline is expressed in a color-differentiated manner, which makes the visual distinction of the pipeline outline higher; the three-dimensional point cloud rendering module makes the image amplitude output by the projection conversion module correspond to the azimuth and distance coordinates, forms three-dimensional point cloud information and performs three-dimensional display; after the above steps, the three-dimensional image output by the microwave three-dimensional imaging system is processed, which can effectively improve the outline display of the pipeline inside the wall, and improve the accuracy of subsequent pipeline detection, identification and positioning. It is a good method worthy of promotion.

[0031] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A microwave imaging three-dimensional image processing method, characterized in that: The following steps are involved: S110, after the microwave 3D imaging system obtains the scattered echo signal from the target, it obtains the slice image of each distance through the 3D image reconstruction algorithm to form a 3D complex image , modulus the 3D complex image to obtain the 3D amplitude image ; S120, for three-dimensional amplitude image The distance dimension Perform amplitude compensation, and the image data after distance compensation is ; S130, using a three-dimensional spatial filter to process the distance-compensated image data of , , Dimensional spatial convolution calculation is performed to obtain the filtered image: ; S140, for the filtered image Perform pixel-by-pixel gamma transformation, and the image after gamma transformation is ; S150, for gamma transformed images Perform logarithmic transformation and obtain the logarithmic transformed image: ; S160, log-transformed image The distance dimension is projected and transformed to obtain the projected image and distance dimension maximum projection coordinates ; S170, using a rainbow color scale mapping table, projects the coordinates of the maximum value of the distance dimension Index the rainbow color scale mapping table to obtain the projected image The color index value corresponding to each pixel point finally obtains the pseudo-color image ; S180, project the image Maximum projection coordinates in the distance dimension The position is filled into the three-dimensional amplitude image to form a three-dimensional point cloud image , the image will be projected Mapped to coordinates The pixel points are used to generate a 3D point cloud image and then sent to the host computer software for 3D display using visualization tools; S190, pseudo color image The pipeline is sent to the pipeline detection, identification and positioning software module for identification and positioning, and the positioning information is sent to the host computer for online display.

2. The microwave imaging three-dimensional image processing method according to claim 1, characterized in that: In S110, the microwave three-dimensional imaging system obtains the scattered echo signal from the target as follows: the frequency band of the signal transmitted by the microwave three-dimensional imaging system is in the range of 3 to 12 GHz, the entire antenna array of the microwave three-dimensional imaging system adopts a sparse array configuration design, and is divided into a sparse transmitting array element and a sparse receiving array element. When the microwave three-dimensional imaging system is triggered by the host computer control timing generator, the sparse transmitting array elements are turned on in time-sharing manner, and the sparse receiving array elements are turned on at the same time to receive the scattered echo signal of the target.

3. The microwave imaging three-dimensional image processing method according to claim 1, characterized in that: In S120, the distance dimension amplitude compensation coefficient used ,in is a three-dimensional amplitude image The coordinates distributed in the distance dimension, is the minimum distance value divided during 3D image reconstruction, It is the maximum value of the distance divided during three-dimensional image reconstruction.

4. The microwave imaging three-dimensional image processing method according to claim 1, characterized in that: In S130, the three-dimensional spatial filter is expressed as , which is used to filter out isolated noise signals in the distance-compensated image data and smooth out the parts where the pixel values ​​fluctuate dramatically. Designed for: ， in, , which are three two-dimensional spatial filters, and finally combined into a three-dimensional spatial filter , spatial convolution calculation includes: three-dimensional spatial filter With image data Convolution is performed pixel by pixel.

5. The microwave imaging three-dimensional image processing method according to claim 1, characterized in that: In S150, the logarithmic transformation process includes: applying a logarithmic transformation with a base of 2 to each pixel, that is, the logarithmic transformation image is .

6. The microwave imaging three-dimensional image processing method according to claim 1, characterized in that: In S170, the rainbow color scale mapping table is represented as ,in Represents the value of the color scale mapping table, Indicates the index value of the mapping table. Indicates the corresponding three-channel color value in the rainbow color mapping table, according to the projected image and distance dimension maximum projection coordinates The values ​​of the three channels corresponding to the rainbow color scale mapping table are ,in express Coordinates in The color value indexed in express Coordinates in The color value indexed in express Coordinates in The color scale value indexed in the middle; Therefore, the pseudo-color floating-point image of the three channels is obtained as follows: ,in , , Respectively expressed as: ； Furthermore, the three-channel floating-point images are converted into 8-bit grayscale images respectively, and the final pseudo-color normalized image is obtained as follows: ,in , , Respectively expressed as: 。

Images