A method, system and apparatus for full focus imaging of a multilayer dielectric structure
By setting a preset spatial coherence selection strategy and a gradient descent iteration strategy, the problems of high computational load and low efficiency in imaging multilayer media structures are solved, and efficient and low-cost all-focus imaging is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ELECTRIC POWER RESEARCH INSTITUTE OF STATE GRID JIBEI ELECTRIC POWER CO LTD
- Filing Date
- 2026-03-25
- Publication Date
- 2026-06-12
AI Technical Summary
Existing multi-layer dielectric structure full-focus imaging technology suffers from high computational load and low efficiency, making it difficult to meet the rapid detection needs in industrial settings. Furthermore, hardware upgrades are costly and offer limited efficiency improvements.
By setting a preset spatial coherence selection strategy and a preset gradient descent iteration strategy, it was determined that the sound propagation time only requires a small number of iterations to converge, and imaging was performed in conjunction with a full matrix data acquisition strategy.
It significantly improves imaging speed while maintaining imaging accuracy, reduces hardware cost requirements, is suitable for conventional computing platforms, and has strong adaptability.
Smart Images
Figure CN122193412A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of defect detection technology for multilayer dielectric structures, and in particular to a method, system and device for full-focus imaging of multilayer dielectric structures. Background Technology
[0002] Multilayer media structural components are widely used in the manufacture of equipment such as connectors in power equipment, composite material components in aerospace, and composite plates in pressure vessels. To ensure equipment safety, it is necessary to detect defects inside them.
[0003] Ultrasonic phased array full-focusing imaging technology offers advantages such as high resolution and global focusing, facilitating the effective detection of internal defects in multi-layered dielectric structures. Current full-focusing imaging schemes require pixel-by-pixel calculation of sound propagation time. However, unlike single-layered dielectric structures, multi-layered dielectric structures exhibit multi-interface refraction characteristics, resulting in complex sound wave propagation paths. Existing methods for calculating sound propagation time employ a global grid exhaustive search based on Fermat's Last Theorem to calculate the optimal refraction point position at each interface. This involves verifying potential refraction point combinations one by one within a large search space across each dielectric interface, with adjacent pixels being solved independently. This leads to high computational load and low efficiency, making it difficult to meet the rapid detection needs of industrial sites. Furthermore, while some solutions exist to improve the efficiency of full-focusing imaging by simplifying the propagation model or increasing hardware computing power, simplifying the propagation model often reduces computational load by ignoring some interface refraction details or simplifying the propagation path, inevitably sacrificing imaging accuracy. Hardware upgrades require replacing the processor with a higher-performance one or adding computing nodes, significantly increasing the purchase and maintenance costs of the detection equipment. Moreover, efficiency improvements are limited by hardware computing power constraints.
[0004] Therefore, how to provide a full-focus imaging scheme for multi-layered dielectric structures that can balance accuracy and efficiency is an urgent problem to be solved. Summary of the Invention
[0005] In view of this, the present invention provides a multi-layer dielectric structure full-focus imaging method, system and device. By setting a preset spatial coherence selection strategy and a preset gradient descent iteration strategy, convergence can be achieved with only a small number of iterations when the target sound propagation time is determined, thereby improving imaging speed while ensuring imaging accuracy.
[0006] To address the aforementioned technical problems, this application provides a method for full-focus imaging of multilayer dielectric structures, comprising: A physical model and coordinate system are established for the M-layer dielectric structure, and the number of dielectric layers in the imaging area to be measured is determined in the physical model for each pixel; M is an integer not less than 2. Each array element under the phased array probe is taken as a target array element and each pixel is taken as a target pixel. The acoustic propagation time relationship between the target array element and the target pixel is determined according to the number of medium layers of the target pixel and the physical model. Based on the position coordinates of the target pixel in the coordinate system and the preset spatial coherence selection strategy, it is determined whether the target pixel has a related pixel corresponding to the target element; If they exist, the target position coordinates of each interface refraction point corresponding to the relevant pixel point are used as the initial values for the position coordinates of each interface refraction point that the sound wave passes through as it propagates from the target array element to the target pixel point. Based on the initial value of the iteration, the sound propagation time relationship, and the preset gradient descent iteration strategy, the target sound propagation time corresponding to the propagation along the optimal propagation path between the target array element and the target pixel, and the target position coordinates of each interface refraction point are determined iteratively. The M-layer medium structure is sampled based on a full matrix data acquisition strategy to obtain full matrix data. Then, based on the target sound propagation time between each array element and each pixel and the full matrix data, the amplitude corresponding to each pixel is determined to obtain the full-focus imaging result of the imaging area.
[0007] Furthermore, the physical model includes first information corresponding to the phased array probe, second information corresponding to each pixel, and basic parameter information of the M-layer dielectric structure. Each of the array elements is disposed on the upper surface of the M-layer dielectric structure; the coordinate system is a rectangular coordinate system established on the upper surface, wherein the x-axis is the lateral direction extending along the array element arrangement direction, and the z-axis is the depth direction.
[0008] Furthermore, the first information includes the position coordinates of each array element in the coordinate system, the total number of array elements, and the array element spacing; the second information includes the horizontal step size, the depth step size, and the position coordinates of each pixel in the coordinate system; the basic parameter information includes the sound velocity and thickness of each medium layer in the M-layer medium structure. Determining the number of media layers in the physical model for each pixel within the imaging region to be measured includes: Based on the thickness of each medium layer under the M-layer medium structure, determine the interface depth coordinates corresponding to each interface in the coordinate system. The pixels are traversed according to a preset scanning order. Based on the position coordinates of the pixels and the depth coordinates of each interface, the number of the medium layer in the physical model corresponding to the pixel is determined.
[0009] Furthermore, based on the number of medium layers of the target pixel and the physical model, the acoustic propagation time relationship between the target array element and the target pixel is determined, including: Based on the number of medium layers of the target pixel, the lateral coordinates corresponding to the interface refraction points through which the sound wave propagates from the target array element to the target pixel are determined as variables to be optimized. Based on Fermat's principle and the physical model, the straight-line lengths corresponding to the propagation paths of the sound wave from the target array element to the target pixel are determined in each layer of the medium. Therefore, the relationship between the sound propagation time between the array element and the target pixel is determined as follows:
[0010] in, This represents the variable to be optimized. Indicates the total propagation time. This represents the straight-line distance that the sound wave travels in the i-th layer of the medium. Represents the sound velocity of the i-th dielectric layer. N This indicates the number of media layers.
[0011] Furthermore, after determining the acoustic propagation time relationship between the target array elements and the target pixels, the process also includes: Determine the gradient analytical expression of the sound propagation time relationship with respect to the lateral coordinates of each of the interface refraction points; The step of iteratively determining the target sound propagation time and the target position coordinates of each interface refraction point when the target array element and the target pixel propagate along the optimal propagation path includes: Based on the sound propagation time relationship and the gradient analytical expression, gradient descent iteration is performed starting from the initial value until a preset termination condition is reached. At this point, the lateral coordinates of each interface refraction point are determined as the target lateral coordinates, and the target sound propagation time is determined according to the target lateral coordinates and the sound propagation time relationship.
[0012] Furthermore, the gradient descent iteration process starting from the initial value includes: In the k-th iteration step, based on the current lateral coordinates of the refraction points at each interface... And the gradient analytical expression, along the negative gradient direction, determine the proposed update value of each of the horizontal coordinates in the (k+1)th iteration step. for:
[0013] in, This indicates that the value corresponding to the gradient determined based on the gradient analytical expression is... gradient value, The preset iteration step size; k is an integer not less than 1; Determined based on the aforementioned sound propagation time relationship. The corresponding total propagation time, and the judgment with Is the corresponding total propagation time less than that of... The corresponding total propagation time; If so, confirm Let k be the updated lateral coordinates of the refraction points at each interface, and let k = k + 1; If not, control Decrease and return the proposed update values for each of the lateral coordinates in the (k+1)th iteration step, determined along the negative gradient direction. The steps.
[0014] Furthermore, the pixels are obtained by dividing the imaging area according to a grid. Each array element under the phased array probe is taken as the target array element, and each pixel is taken as the target pixel, including: Each element of the phased array probe is taken as a target element, and the pixels under the grid are traversed row by row for the target element to determine the corresponding target pixels. Based on the position coordinates of the target pixel in the coordinate system and a preset spatial coherence selection strategy, it is determined that the target pixel has related pixels corresponding to the target element, including: Based on the position coordinates of the target pixel in the coordinate system, determine whether the target pixel is the first pixel in the row where the target pixel is located; If it is not the first pixel, when it is determined that the left neighbor pixel of the target pixel in the row has the target horizontal coordinate corresponding to each interface refraction point, the left neighbor pixel is determined as the relevant pixel; when it is determined that the left neighbor pixel has not the target horizontal coordinate corresponding to each interface refraction point, the upper neighbor pixel of the target pixel in the column is determined as the relevant pixel.
[0015] Furthermore, when determining that the target pixel is the first pixel in the row containing the target pixel, the process includes: Determine the lateral coordinates of the intersection point of the target array element and the target pixel with the nth interface. for:
[0016] Where 1≤n≤N-1, and N is the number of dielectric layers; Indicates the lateral coordinates of the target array element, ( , () represents the position coordinates of the target pixel. Represents the interface depth coordinates of the nth interface; Each of the above The initial values of the lateral coordinates of each interface refraction point along the path of the sound wave as it propagates from the target array element to the target pixel are used as the iterative initial values.
[0017] To address the aforementioned technical problems, the present invention also provides a multi-layer dielectric structure full-focus imaging system, comprising: The module is used to establish a physical model and coordinate system for the M-layer dielectric structure, and to determine the number of dielectric layers in the physical model for each pixel in the imaging area to be measured; M is an integer not less than 2. The sound propagation time determination module takes each array element under the phased array probe as the target array element and each pixel as the target pixel, and determines the sound propagation time relationship between the target array element and the target pixel based on the number of medium layers of the target pixel and the physical model. The judgment module is used to determine whether there are related pixels corresponding to the target element based on the position coordinates of the target pixel in the coordinate system and a preset spatial coherence selection strategy; if so, it proceeds to the initial value determination module. The initial value determination module is used to take the target position coordinates of each interface refraction point corresponding to the relevant pixel as the iterative initial value of the position coordinates of each interface refraction point that the sound wave passes through as it propagates from the target array element to the target pixel. The iteration module is used to iteratively determine the target sound propagation time and the target position coordinates of each interface refraction point when the sound propagates along the optimal propagation path between the target array element and the target pixel, based on the initial iteration value, the sound propagation time relationship and the preset gradient descent iteration strategy. The sampling and imaging module is used to sample the M-layer medium structure based on a full matrix data acquisition strategy to obtain full matrix data. Then, based on the target sound propagation time between each array element and each pixel and the full matrix data, the amplitude corresponding to each pixel is determined to obtain the full-focus imaging result of the imaging area.
[0018] To address the aforementioned technical problems, the present invention also provides a multi-layer dielectric structure full-focus imaging device, comprising: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the steps of the multilayer dielectric structure total focusing imaging method as described above.
[0019] The beneficial effects of this invention are as follows: This application provides a method, system, and apparatus for full-focus imaging of multilayer dielectric structures. The method includes establishing a physical model and coordinate system for the M-layer dielectric structure, and determining the number of dielectric layers in the physical model for each pixel in the imaging area to be measured; taking each element of the phased array probe as a target element and each pixel as a target pixel, and determining the acoustic propagation time relationship between the target element and the target pixel based on the number of dielectric layers and the physical model; determining whether there are related pixels corresponding to the target element based on the position coordinates of the target pixel in the coordinate system and a preset spatial coherence selection strategy; if so, selecting the related pixels... The target position coordinates of each interface refraction point corresponding to the point are used as the initial values for the iterative calculation of the position coordinates of each interface refraction point along which the sound wave propagates from the target array element to the target pixel. Based on the initial values, the sound propagation time relationship, and the preset gradient descent iterative strategy, the target sound propagation time and the target position coordinates of each interface refraction point are determined iteratively when the sound propagates along the optimal propagation path between the target array element and the target pixel. Based on the full matrix data acquisition strategy, the M-layer medium structure is sampled to obtain full matrix data. Then, based on the target sound propagation time between each array element and each pixel and the full matrix data, the amplitude corresponding to each pixel is determined to obtain the full-focus imaging result of the imaging area. As can be seen, this scheme, by setting a preset spatial coherence selection strategy, can transfer the target position coordinates of each interface refraction point corresponding to the relevant pixels, which have been iteratively converged, to the current target pixel as the initial value for iteration. Based on the preset gradient descent iteration strategy, it can converge with only a small number of iterations when determining the sound propagation time of the target, which is beneficial to improving the imaging speed. Moreover, this scheme does not require simplification of the propagation process or refraction details, which is beneficial to ensuring imaging accuracy and improving imaging efficiency while ensuring imaging accuracy. In addition, this scheme does not rely on high-computing hardware and can achieve high-precision and fast imaging on conventional computing platforms, which is beneficial to reducing costs and is conducive to practical applications in industrial fields.
[0020] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application, it can be implemented according to the contents of the specification. In order to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description
[0021] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart of a multi-layer dielectric structure full-focus imaging method provided by the present invention; Figure 2A schematic diagram of sound wave transmission in a four-layer dielectric structure provided by the present invention; Figure 3 This is a schematic diagram of a full-focus imaging result of a four-layer dielectric structure obtained based on the technical solution provided in this application, as provided by the present invention. Figure 4 This is a schematic diagram of a full-focus imaging result of a four-layer dielectric structure obtained based on a global grid exhaustive search method. Figure 5 A schematic diagram of a multi-layer dielectric structure full-focus imaging system provided by the present invention; Figure 6 This is a schematic diagram of a multi-layer dielectric structure for a total focusing imaging device provided by the present invention. Detailed Implementation
[0022] The core of this invention is to provide a multi-layer dielectric structure full-focus imaging method, system, and device. By setting a preset spatial coherence selection strategy and a preset gradient descent iteration strategy, convergence can be achieved with only a small number of iterations when the target sound propagation time is determined, thereby improving imaging speed while ensuring imaging accuracy.
[0023] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.
[0024] The terms "first," "second," etc., used in this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class, without limiting the number of objects; for example, a first object can be one or more. Furthermore, in this application, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects have an "or" relationship.
[0025] Please refer to Figure 1 , Figure 1 The flowchart illustrates a multi-layer dielectric structure full-focus imaging method provided by this invention.
[0026] The full-focus imaging method for multilayer dielectric structures includes: S11: Establish a physical model and coordinate system for the M-layer dielectric structure, and determine the number of dielectric layers in the physical model for each pixel in the imaging area to be measured; M is an integer not less than 2. S12: Take each element of the phased array probe as the target element and each pixel as the target pixel. Determine the acoustic propagation time relationship between the target element and the target pixel based on the number of medium layers and the physical model of the target pixel. S13: Based on the position coordinates of the target pixel in the coordinate system and the preset spatial coherence selection strategy, determine whether there are related pixels corresponding to the target element; if so, proceed to S14. S14: Use the target position coordinates of each interface refraction point corresponding to the relevant pixel as the initial value of the position coordinates of each interface refraction point that the sound wave passes through as it propagates from the target array element to the target pixel. S15: Based on the initial value of the iteration, the relationship between the sound propagation time and the preset gradient descent iteration strategy, iteratively determine the target sound propagation time and the target position coordinates of each interface refraction point when the target array element and the target pixel point propagate along the optimal propagation path. S16: Based on the full matrix data acquisition strategy, the M-layer medium structure is sampled to obtain full matrix data. Then, based on the target sound propagation time between each array element and each pixel and the full matrix data, the amplitude corresponding to each pixel is determined to obtain the full-focus imaging result of the imaging area.
[0027] In this embodiment, the scheme can be applied to imaging scenarios of any multi-layer dielectric structure with two or more layers, and it has good adaptability to combinations of media with different sound velocities and thicknesses. The multi-layer dielectric structure can be various modules, devices, or equipment with multi-layer dielectric structures. The imaging area refers to any imaging area to be tested under the M-layer dielectric structure. After obtaining the corresponding full-focus imaging result, it can be determined whether there is an anomaly at the imaging area based on the result and the preset anomaly judgment condition, so as to realize the internal defect detection of the multi-layer dielectric structure. The preset anomaly judgment condition can be flexibly set according to the actual application needs, and is not particularly limited here.
[0028] To elaborate, regarding step S11, each array element under the phased array probe is positioned on the upper surface of the M-layer dielectric structure. That is, each array element is arranged sequentially on this upper surface in the same row corresponding to the imaging area. Here, the upper surface can be understood as the incident surface of the sound waves; please refer to... Figure 2 , Figure 2 This is a schematic diagram of sound wave transmission in a four-layer dielectric structure provided by the present invention. It can be seen that the imaging region is essentially a two-dimensional plane under the M-layer dielectric structure, and the array elements are arranged sequentially in the lateral direction of this two-dimensional plane. Figure 2The diagram illustrates the propagation path of sound waves when sampling according to the full matrix data acquisition strategy. The sound waves start from the array element ZY1, which is the transmitter, and are refracted through each medium layer to reach the pixel XS1. They are then refracted through each medium layer and transmitted back to the array element ZY2, which is one of the receivers.
[0029] The coordinate system is a Cartesian coordinate system established on this upper surface, where the horizontal direction is defined as the x-axis, that is, the horizontal direction extending along the array element arrangement direction, and the depth direction is defined as the z-axis, such as... Figure 2 As shown; because each element of the phased array probe is set close to the upper surface, all elements are close to the z=0 plane, that is, the depth coordinate of each element is 0. e The horizontal coordinates of each array element can be represented as:
[0030] in, e Indicates the index of the array element. This represents the horizontal coordinate of the array element used as a reference. d This indicates the element spacing included in the first information under the physical model. ，N e This indicates the total number of array elements included in the first information.
[0031] The established physical model includes first information corresponding to the phased array probe, second information corresponding to each pixel, and basic parameter information of the M-layer dielectric structure. The basic parameter information includes the thickness of each dielectric layer. Therefore, based on the thickness of each dielectric layer under the M-layer dielectric structure, the interface depth coordinates corresponding to each interface in the coordinate system can be determined. Here, the interface refers to the junction between two adjacent dielectric layers, corresponding to... Figure 2 As shown, taking M=4 as an example, the layers are sequentially labeled as layer 1 to layer 4 along the depth direction. Layer 1 is the water layer, layer 2 is the copper layer, layer 3 is the tin layer, and layer 4 is the aluminum layer. The first interface is the junction between the water layer and the copper layer, the second interface is the junction between the copper layer and the tin layer, and the third interface is the junction between the tin layer and the aluminum layer. For example, assuming the thicknesses of each layer from layer 1 to layer 4 are sequentially... h 1 、h 2 、h 3 、h 4. The interface depth coordinates of the first interface The interface depth coordinates of the second interface The interface depth coordinates of the third interface are .
[0032] It is understandable that the position coordinates of each interface refraction point along the path of the sound wave from the target array element to the target pixel include the lateral coordinates and the depth coordinates. Since the depth coordinates are the interface depth coordinates of the interface where the interface refraction point is located, in practical applications, we can focus only on determining the lateral coordinates of each interface refraction point. That is, the iterative determination of the position coordinates of each interface refraction point in this application can be simplified to the iterative determination of the lateral coordinates of each interface refraction point.
[0033] The imaging region is divided into a grid to obtain individual pixels. Specifically, the imaging region is divided into a grid within its horizontal and depth dimensions, with a horizontal step size of Δ. x The depth step size is Δ z Each node of the grid corresponds to a pixel.
[0034] Furthermore, according to step S12, the sound propagation time expression between the target array element and the target pixel is determined. This expression is essentially an expression with the total propagation time as the value to be optimized and the lateral coordinates of each interface refraction point as the variable to be optimized. Taking advantage of the high similarity of the optimal propagation paths of adjacent pixels, a preset spatial coherence selection strategy is set to determine the relevant pixels. Then, according to steps S13 and S14, the target position coordinates of each interface refraction point, which have been iteratively solved for the relevant pixels, are used as the initial values for the iteration of the lateral coordinates of each interface refraction point of the target pixel. It should be understood that the target position coordinates are the target position coordinates of each interface refraction point corresponding to the propagation along the optimal propagation path between the target array element and the adjacent pixels.
[0035] Step S15 aims to use the initial iteration value as a starting point to perform gradient descent iterative correction on the position coordinates of each interface refraction point of the target pixel to find the optimal propagation path that minimizes the total propagation time of the sound wave from the target array element to the target pixel. When a preset termination condition is reached, the position coordinates of each interface refraction point are determined as the target position coordinates, and the total propagation time obtained by substituting the target position coordinates into the sound propagation time relationship is the target sound propagation time. It can be understood that for each array element, the target sound propagation time between it and each pixel is determined, and this step is repeated for each array element until the target sound propagation time between each array element and each pixel is obtained, generating a complete three-dimensional propagation time matrix. OWT ( x p , z p , e ),in,( x p , z p () is used to represent the position coordinates of a pixel in a coordinate system.
[0036] For step S16, the step of sampling the M-layer dielectric structure based on the full matrix data acquisition strategy to obtain full matrix data includes: adopting a full matrix data acquisition mode, sampling each array element under the phased array probe as a sound wave transmitting array element in sequence, and sampling all other array elements as sound wave receiving array elements to obtain the original full matrix data. This original full matrix data is essentially... N e × N e Group channel data; more specifically, considering that the original full matrix data usually contains DC components and system noise, the original full matrix data can be preprocessed, such as uniformly removing DC components and bandpass filtering, to obtain full matrix data for subsequent full-focus imaging.
[0037] Furthermore, based on the fundamental principles of total focusing imaging, and based on the aforementioned full matrix data and OWT ( x p , z p , e ) performs time-delayed superposition imaging; more specifically, for each pixel point within the imaging area ( x p , z p ), traverse all launch array elements Tx ( x Take 1 to N e ) and receiving array elements Rx ( x Take 1 to N e ), that is, from OWT ( x p , z p , e Extract the corresponding pixel and the emission array element. Tx One-way sound propagation time between launch targets t Tx ( x p , z p ), corresponding to the pixel and the receiving array element Rx One-way target sound propagation time between the receivers t Rx ( x p , z p This allows us to obtain the total sound wave propagation path time.t sum for:
[0038] By combining and summing the corresponding full matrix data acquisition results, the pixel points can be obtained. x p , z p amplitude I ( x p , z p )for:
[0039] Then, by utilizing the amplitude of each pixel, a fully focused image of the imaging area is generated. The M-layer dielectric structure can be divided into multiple imaging areas to be tested. By performing the above steps on each imaging area, a complete fully focused image of the M-layer dielectric structure can be obtained.
[0040] In summary, this application provides a full-focus imaging method for multi-layer dielectric structures. Unlike existing exhaustive search imaging schemes, this method, by setting a preset spatial coherence selection strategy, can transfer the target position coordinates of each interface refraction point corresponding to relevant pixels, which have been iteratively converged, to the current target pixel as the initial value for iteration. Based on a preset gradient descent iteration strategy, it can converge with only a small number of iterations when determining the sound propagation time of the target, which is beneficial to improving the imaging speed. Moreover, this method does not require simplification of the propagation process or refraction details, which is beneficial to ensuring imaging accuracy and improving imaging efficiency while maintaining imaging accuracy. In addition, this method does not rely on high-computing hardware and can achieve high-precision and fast imaging on conventional computing platforms, which is beneficial to reducing costs and practical applications in industrial settings.
[0041] Based on the above embodiments: In some embodiments, the first information includes the position coordinates of each element in the coordinate system, the total number of elements, and the spacing between elements; the second information includes the horizontal step size, the depth step size, and the position coordinates of each pixel in the coordinate system; the basic parameter information includes the sound velocity and thickness of each layer of the medium in the M-layer medium structure. Determine the number of media layers in the physical model for each pixel in the imaging region to be tested, including: Based on the thickness of each medium layer under the M-layer medium structure, determine the interface depth coordinates corresponding to each interface in the coordinate system; The system traverses each pixel in a preset scanning order and determines the number of medium layers in the physical model corresponding to each pixel based on the pixel's position coordinates and the depth coordinates of each interface.
[0042] It should also be noted that the horizontal step size here essentially represents the horizontal distance between two adjacent pixels in the same row within the grid of the imaging area, while the depth step size essentially represents the vertical distance between two adjacent pixels in the same column within the same grid. The preset scanning order can be either row-by-row or column-by-column scanning; no particular limitation is made here. As can be seen, the above settings can ensure that the established physical model can accurately characterize the physical characteristics of the detection scene corresponding to the imaging area.
[0043] In some embodiments, determining the acoustic propagation time relationship between the target array element and the target pixel based on the number of medium layers and the physical model of the target pixel includes: Based on the number of media layers of the target pixel, the lateral coordinates corresponding to the interface refraction points through which the sound wave propagates from the target array element to the target pixel are determined as variables to be optimized. Based on Fermat's principle, the straight-line lengths corresponding to the propagation path of sound waves from the target array element to the target pixel are determined according to the physical model, and thus the relationship between the sound propagation time between the array element and the target pixel is determined as follows:
[0044] in, Represents the variable to be optimized. Indicates the total propagation time. This represents the straight-line distance that a sound wave travels in the i-th layer of medium. Represents the sound velocity of the i-th dielectric layer. N Indicates the number of dielectric layers.
[0045] In this embodiment, corresponding to the combination of target array elements and target pixels, an expression for the sound propagation time of the sound wave starting from the upper surface where the target array element is located, passing through multiple interfaces for refraction, and finally reaching the target pixel is established. Specifically, according to Fermat's principle, the actual sound wave path has the shortest propagation time. In the layered isotropic medium considered in this application, the propagation path within each medium layer is approximately a straight line. Therefore, the entire propagation path from the upper surface to the target pixel can be decomposed into several straight line segments.
[0046] If the target pixel is located in the first layer of medium, i.e., N=1, then the straight-line length between the target array element and the target pixel is directly determined as the straight-line distance during sound wave propagation, and the ratio of the straight-line length to the sound speed in the first layer of medium is the total propagation time.
[0047] If the target pixel is not located in the first layer of the medium (i.e., 2≤N≤M), then for a target pixel located in the Nth layer of the medium, there are N paths (corresponding to N straight-line distances) from the upper surface of the target element to the target pixel, passing through layers 1 to N, and passing through N-1 interfaces. The lateral coordinates of the unknown, unoptimized interface refraction point are defined on the nth interface. For 1≤n≤N-1, the variable to be optimized, which consists of the lateral coordinates of the interface refraction points along the path of the sound wave from the target array element to the target pixel, can be represented in vector form as follows: .
[0048] Given a certain set In this case, based on the physical model and geometric relationships, the straight-line distance corresponding to the propagation of the sound wave in the i-th medium layer, i.e., the sound path, can be calculated. For example, the straight-line distance corresponding to the propagation of the sound wave in the 1st medium layer can be determined based on the position coordinates of the target array element and ( , The calculation shows that the straight-line distance for the sound wave to propagate in the second medium layer can be determined by ( ). , )and( , The calculation is as follows; since the depth coordinates of the refraction points at each interface are known, it is only necessary to iteratively optimize the lateral coordinates of the refraction points at each interface to obtain the above sound propagation time relationship.
[0049] For example, corresponding to Figure 2 The target pixel XS1 is located in the 4th layer of the medium, and the position coordinates of the target element ZY1 are ( The coordinates from (0) to the target pixel XS1 are ( x p , z p The propagation path of ) includes four segments, namely from ( The position coordinates of the first interface refraction point (0) , ),from( , The coordinates of the position of the refraction point from the second interface ( , ),from( , The coordinates of the position of the refraction point from the third interface ( , ),from( , The position coordinates from the target pixel XS1 () x p , zp The four propagation paths are as follows: , , , The sound propagation time expression at this point is: .
[0050] As can be seen, the above settings reliably and accurately express the problem of determining the acoustic propagation time between the target array elements and the target pixels through the acoustic propagation time relationship, laying the foundation for subsequent gradient descent iterations.
[0051] In some embodiments, after determining the acoustic propagation time relationship between the target array element and the target pixel, the method further includes: Determine the gradient analytical expression of the sound propagation time relationship with respect to the transverse coordinates of the refraction points at each interface; The steps for iteratively determining the target acoustic propagation time and the target position coordinates of each interface refraction point during the propagation along the optimal propagation path between the target array element and the target pixel include: Based on the sound propagation time relationship and gradient analytical expression, gradient descent iteration is performed starting from the initial value until the preset termination condition is reached. At this time, the lateral coordinates of the refraction points of each interface are determined as the target lateral coordinates, and the target sound propagation time is determined according to the target lateral coordinates and the sound propagation time relationship.
[0052] First, it should be noted that determining the gradient analytical expression of the sound propagation time relationship with respect to the transverse coordinates of the refraction points at each interface is essentially a... Calculate the first-order partial derivatives for the lateral coordinates of the refraction points at each interface, since each straight path segment... Euclidean distance is used for calculation, which has the form of the sum of squares and the square root, therefore... Finding partial derivatives is analytically feasible.
[0053] For example, when the target pixel is located in the second layer of the medium, the variable to be optimized is only The corresponding sound propagation time relationship is:
[0054] Ask about it The first-order partial derivatives of can be used to obtain the analytical expression for the gradient at this point:
[0055] When the target pixel is located in the Nth layer of the medium and N is greater than 2, the gradient analytical expression is: ;in, , , .
[0056] For example, when the target pixel is located in the third layer of the medium, the variable to be optimized is: The corresponding sound propagation time relationship is:
[0057] The corresponding gradient analytical expression is:
[0058] in, , ; When the target pixel is located in the 4th layer of the medium, the variable to be optimized is: The corresponding sound propagation time relationship is:
[0059] The corresponding gradient analytical expression is:
[0060] in , , .
[0061] Using this gradient analytical expression, starting from the initial iteration value, and combining it with the sound propagation time relationship, the position coordinates of each interface refraction point of the target pixel are jointly corrected by gradient descent iteratively until a preset termination condition is reached. At this point, the position coordinates of each interface refraction point are determined to be the optimal target position coordinates, and the total propagation time obtained by substituting these target position coordinates into the sound propagation time relationship is the shortest target sound propagation time. It should be noted that the joint gradient descent iterative correction of the position coordinates of multiple interface refraction points helps ensure that all interface refraction points simultaneously satisfy the refraction law and Fermat's principle of their respective interfaces, guaranteeing the accuracy of the multi-layer refraction path. The preset termination condition can be set as follows: the gradient vector norm is less than a preset gradient threshold; the change in the position coordinates of each interface refraction point between two adjacent iterations is less than a preset position threshold; the difference in the total propagation time between two adjacent iterations is less than a preset error threshold; or the number of iterations reaches a preset number of iterations. No specific limitations are imposed here.
[0062] As can be seen, the above method can accurately guide the iteration to converge efficiently to the optimal solution based on the gradient analytical expression of the target pixel points located in different media layers.
[0063] In some embodiments, the gradient descent iteration process starting from the initial value includes: In the k-th iteration step, based on the current lateral coordinates of the refraction points at each interface... And the gradient analytical expression, along the negative gradient direction, determine the proposed update values of each horizontal coordinate in the (k+1)th iteration step. for:
[0064] in, This indicates that the expression corresponding to the gradient is determined based on the analytical expression. gradient value, The preset iteration step size; k is an integer not less than 1; Determined based on the time-propagation relationship of sound The corresponding total propagation time, and the judgment with Is the corresponding total propagation time less than that of... The corresponding total propagation time; If so, confirm Let k be the updated lateral coordinates of the refraction points at each interface, and let k = k + 1; If not, control Decrease and return the proposed update values for each lateral coordinate in the (k+1)th iteration step, determined along the negative gradient direction. The steps.
[0065] Specifically, Essentially, it is determined by the gradient analytical expression that corresponds to The vector formed by the gradient values, The specific values can be flexibly set according to the actual application, and no special limitation is made here; in order to balance convergence speed with solution stability and accuracy, this application sets the above-mentioned gradient descent iteration method with adaptive adjustment of iteration step size to determine the optimal propagation path that minimizes the total propagation time, that is, when it is determined that The corresponding total propagation time is less than When the corresponding total propagation time is reached, it indicates that the total propagation time has successfully decreased, and the update is received; when determining... The corresponding total propagation time is not less than When the corresponding total propagation time is reached, it indicates that the current iteration step size is too large, causing it to exceed the minimum point. Therefore, the update is rejected, and control is implemented. Reduce to retry iterative updates; control here One way to achieve adaptive reduction is to use a pre-set step size decay coefficient. γ renew The updated It returns the proposed update values for each lateral coordinate in the (k+1)th iteration step, determined along the negative gradient direction. The steps continue until the total propagation time corresponding to the new proposed update value is less than [a certain value]. The corresponding total propagation time; here 0 < γ <1, for example, γ=0.5; it can be seen that this setting can effectively avoid iteration oscillation or divergence caused by a fixed iteration step size.
[0066] It should be noted that in the next iteration, the initial values will still be prioritized. Determine at this time When control is needed The reduction method described above will be executed again when the value is reduced.
[0067] It should also be noted that in the k-th iteration step, we can first determine the current... The system determines whether to stop iteration based on a preset termination condition; if so, it determines the current iteration. Let the target lateral coordinates of each interface refraction point be denoted; otherwise, proceed to the step of determining the proposed update values of each lateral coordinate in the (k+1)th iteration along the negative gradient direction. The steps are as follows; it is understandable that different preset termination conditions can be used to determine whether to terminate the iteration, and no special restrictions are made here.
[0068] In some embodiments, pixels are obtained by dividing the imaging area according to a grid. Each element of the phased array probe is taken as the target element, and each pixel is taken as the target pixel, including: Each element under the phased array probe is taken as the target element, and the pixels under the grid are traversed row by row for the target element to determine the corresponding target pixel. Based on the position coordinates of the target pixel in the coordinate system and the preset spatial coherence selection strategy, it is determined that the target pixel has related pixels corresponding to the target array element, including: Based on the position coordinates of the target pixel in the coordinate system, determine whether the target pixel is the first pixel in the row where the target pixel is located; If it is not the first pixel, when the left neighbor pixel of the target pixel in the row has been determined to correspond to the target horizontal coordinates of each interface refraction point, the left neighbor pixel is determined to be the relevant pixel; when the left neighbor pixel has not been determined to correspond to the target horizontal coordinates of each interface refraction point, the upper neighbor pixel of the target pixel in the column is determined to be the relevant pixel.
[0069] In this embodiment, the selection and initial value determination of related pixels based on spatial coherence is one of the key steps in effectively reducing the workload of gradient descent iteration and improving the overall imaging speed determination efficiency. Specifically, based on the refined grid delineation of the imaging area, the optimal transmission path between adjacent pixels changes continuously and smoothly in space. The corresponding acoustic propagation path and the position coordinates of the interface refraction point are highly similar, meaning that the position coordinates of the interface refraction point corresponding to adjacent pixels will not jump. Therefore, by taking each array element as the target array element and traversing each pixel row by row for the target array element to determine the corresponding target pixel, the pixel with the target acoustic propagation time and target position coordinates has been determined. Its corresponding target position coordinates can be used as the initial value for the position coordinates of its neighboring pixels, thereby avoiding a global search and reducing the convergence difficulty of subsequent iterations.
[0070] Specifically, when the target pixel is not the first pixel in its row, it can be determined as the relevant pixel when the left neighbor pixel in the row has been determined to correspond to the target's horizontal coordinates at each interface refraction point. Here, the left neighbor pixel refers to the target pixel. x p , z p The pixel corresponding to one horizontal step to the left of its current row. x p-Δx , z p Then the initial value of the horizontal coordinate of the target pixel is... The target horizontal coordinates corresponding to the left neighboring pixel. Considering that in some situations, such as changes in the scanning strategy, the target's horizontal coordinates corresponding to the left neighboring pixel may not yet be calculated, when it is determined that the target's horizontal coordinates corresponding to each interface refraction point are not determined for the left neighboring pixel, the upper neighboring pixel in the column of the target pixel is determined as the relevant pixel. Here, the upper neighboring pixel refers to the target pixel. x p , z p The pixel corresponding to a depth step above its column. x p , z p-Δz Then the initial value of the horizontal coordinate of the target pixel is... The target horizontal coordinates corresponding to the upper neighbor pixel. .
[0071] In some embodiments, determining that the target pixel is the first pixel in the row containing the target pixel includes: Determine the horizontal coordinates of the intersection point of the line connecting the target array element and the target pixel with the nth interface. for:
[0072] Where 1≤n≤N-1, and N is the number of dielectric layers; Indicates the horizontal coordinate of the target array element, ( , () represents the position coordinates of the target pixel. Represents the interface depth coordinates of the nth interface; Each The initial values of the lateral coordinates of the refraction points at each interface along the path of the sound wave as it propagates from the target array element to the target pixel are used for iteration.
[0073] Specifically, when it is determined that the target pixel is the first pixel in the row where the target pixel is located, it also means that there is no related pixel corresponding to the target element. Then, the initial value of the iteration can be determined in the above manner, and the process can proceed to step S15 for iteration.
[0074] To verify the effectiveness of the technical solution provided in this application, simulation experiments were conducted based on a typical four-layer dielectric structure, such as... Figure 2 As shown, the thickness of the water layer was set to 10 mm, the copper layer to 5 mm, the tin layer to 3 mm, and the aluminum layer to 12 mm. An image of a 20 mm × 30 mm area was created using the same computing platform (Intel(R) i7-14650HX 2.20 GHz, MATLAB environment) and imaging resolution (0.1 mm). Please refer to... Figure 3 and Figure 4 , Figure 3 This is a schematic diagram of a full-focus imaging result of a four-layer dielectric structure obtained based on the technical solution provided in this application, as provided by the present invention. Figure 4 This is a schematic diagram of the full-focus imaging result of a four-layer dielectric structure obtained based on a global grid exhaustive search method. The traditional global grid exhaustive search method requires a high-density traversal of discrete points for each interface, which takes 276.47 seconds. In contrast, the scheme of this application only takes 3.31 seconds. The results show that the imaging efficiency of the scheme of this application is improved by nearly 80 times.
[0075] To assess whether the imaging quality of the technical solution provided in this application is compromised, the peak signal-to-noise ratio (PSNR) is introduced as an evaluation metric for calculation. The PSNR of the imaging result of the technical solution provided in this application is 57.18 dB compared with the traditional global grid exhaustive search method. This indicates that the technical solution provided in this application significantly improves computational efficiency while achieving image quality comparable to the traditional global grid exhaustive search method, thus demonstrating the effectiveness of the technical solution provided in this application.
[0076] Please refer to Figure 5 , Figure 5 This is a schematic diagram of a multi-layer dielectric structure full-focus imaging system provided by the present invention.
[0077] This multi-layer dielectric structure full-focus imaging system includes: Module 21 is used to establish a physical model and coordinate system for the M-layer dielectric structure and to determine the number of dielectric layers in the physical model for each pixel in the imaging area to be measured; M is an integer not less than 2. The sound propagation time determination module 22 takes each array element under the phased array probe as the target array element and each pixel as the target pixel, and determines the sound propagation time relationship between the target array element and the target pixel based on the number of medium layers and the physical model of the target pixel. Judgment module 23 is used to determine whether there are related pixels corresponding to the target element based on the position coordinates of the target pixel in the coordinate system and the preset spatial coherence selection strategy; if there are, it enters the initial value determination module 24. The initial value determination module 24 is used to take the target position coordinates of each interface refraction point corresponding to the relevant pixel as the iterative initial value of the position coordinates of each interface refraction point that the sound wave passes through from the target array element to the target pixel. Iteration module 25 is used to iteratively determine the target sound propagation time and the target position coordinates of each interface refraction point when the sound propagates along the optimal propagation path between the target array element and the target pixel, based on the initial iteration value, the sound propagation time relationship and the preset gradient descent iteration strategy. The sampling and imaging module 26 is used to sample the M-layer medium structure based on the full matrix data acquisition strategy to obtain full matrix data. Then, based on the target sound propagation time between each array element and each pixel and the full matrix data, the amplitude corresponding to each pixel is determined to obtain the full-focus imaging result of the imaging area.
[0078] For a description of the multilayer dielectric structure full-focus imaging system provided in this application, please refer to the above-described embodiments of the multilayer dielectric structure full-focus imaging method, which will not be repeated here.
[0079] Please refer to Figure 6 , Figure 6 This is a schematic diagram of a multi-layer dielectric structure for a total focusing imaging device provided by the present invention.
[0080] The multilayer dielectric structure full-focus imaging device includes: Memory 31 is used to store computer programs; The processor 32 is used to execute the computer program to implement the steps of the multilayer dielectric structure full-focus imaging method as described above.
[0081] For a description of the multilayer dielectric structure full-focus imaging device provided in this application, please refer to the above-described embodiments of the multilayer dielectric structure full-focus imaging method, which will not be repeated here.
[0082] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section. Relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0083] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for full-focus imaging of multilayer dielectric structures, characterized in that, include: A physical model and coordinate system are established for the M-layer dielectric structure, and the number of dielectric layers in the imaging area to be measured is determined in the physical model for each pixel. M is an integer not less than 2; Each array element under the phased array probe is taken as a target array element and each pixel is taken as a target pixel. The acoustic propagation time relationship between the target array element and the target pixel is determined according to the number of medium layers of the target pixel and the physical model. Based on the position coordinates of the target pixel in the coordinate system and the preset spatial coherence selection strategy, it is determined whether the target pixel has a related pixel corresponding to the target element; If they exist, the target position coordinates of each interface refraction point corresponding to the relevant pixel point are used as the initial values for the position coordinates of each interface refraction point that the sound wave passes through as it propagates from the target array element to the target pixel point. Based on the initial value of the iteration, the sound propagation time relationship, and the preset gradient descent iteration strategy, the target sound propagation time corresponding to the propagation along the optimal propagation path between the target array element and the target pixel, and the target position coordinates of each interface refraction point are determined iteratively. The M-layer medium structure is sampled based on a full matrix data acquisition strategy to obtain full matrix data. Then, based on the target sound propagation time between each array element and each pixel and the full matrix data, the amplitude corresponding to each pixel is determined to obtain the full-focus imaging result of the imaging area.
2. The multi-layer dielectric structure full-focus imaging method as described in claim 1, characterized in that, The physical model includes first information corresponding to the phased array probe, second information corresponding to each pixel, and basic parameter information of the M-layer dielectric structure. Each of the array elements is disposed on the upper surface of the M-layer dielectric structure; the coordinate system is a rectangular coordinate system established on the upper surface, wherein the x-axis is the lateral direction extending along the array element arrangement direction, and the z-axis is the depth direction.
3. The multi-layer dielectric structure full-focus imaging method as described in claim 2, characterized in that, The first information includes the position coordinates of each array element in the coordinate system, the total number of array elements, and the spacing between array elements; the second information includes the horizontal step size, the depth step size, and the position coordinates of each pixel in the coordinate system; the basic parameter information includes the sound velocity and thickness of each medium layer in the M-layer medium structure. Determining the number of media layers in the physical model for each pixel within the imaging region to be measured includes: Based on the thickness of each medium layer under the M-layer medium structure, determine the interface depth coordinates corresponding to each interface in the coordinate system. The pixels are traversed according to a preset scanning order. Based on the position coordinates of the pixels and the depth coordinates of each interface, the number of the medium layer in the physical model corresponding to the pixel is determined.
4. The multi-layer dielectric structure full-focus imaging method as described in claim 2, characterized in that, Based on the number of medium layers of the target pixel and the physical model, the acoustic propagation time relationship between the target array element and the target pixel is determined, including: Based on the number of medium layers of the target pixel, the lateral coordinates corresponding to the interface refraction points through which the sound wave propagates from the target array element to the target pixel are determined as variables to be optimized. Based on Fermat's principle and the physical model, the straight-line lengths corresponding to the propagation paths of the sound wave from the target array element to the target pixel are determined in each layer of the medium. Therefore, the relationship between the sound propagation time between the array element and the target pixel is determined as follows: in, This represents the variable to be optimized. Indicates the total propagation time. This represents the straight-line distance that the sound wave travels in the i-th layer of medium. Represents the sound velocity of the i-th dielectric layer. N This indicates the number of media layers.
5. The multi-layer dielectric structure full-focus imaging method as described in claim 2, characterized in that, After determining the acoustic propagation time relationship between the target array element and the target pixel, the method further includes: Determine the gradient analytical expression of the sound propagation time relationship with respect to the lateral coordinates of each of the interface refraction points; The step of iteratively determining the target sound propagation time and the target position coordinates of each interface refraction point when the target array element and the target pixel propagate along the optimal propagation path includes: Based on the sound propagation time relationship and the gradient analytical expression, gradient descent iteration is performed starting from the initial value until a preset termination condition is reached. At this point, the lateral coordinates of each interface refraction point are determined as the target lateral coordinates, and the target sound propagation time is determined according to the target lateral coordinates and the sound propagation time relationship.
6. The multi-layer dielectric structure full-focus imaging method as described in claim 5, characterized in that, The process of performing gradient descent iterations starting from the initial value includes: In the k-th iteration step, based on the current lateral coordinates of the refraction points at each interface... And the gradient analytical expression, along the negative gradient direction, determine the proposed update value of each of the horizontal coordinates in the (k+1)th iteration step. for: in, This indicates that the value corresponding to the gradient determined based on the gradient analytical expression is... gradient value, The preset iteration step size; k is an integer not less than 1; Determined based on the aforementioned sound propagation time relationship. The corresponding total propagation time, and the judgment with Is the corresponding total propagation time less than that of... The corresponding total propagation time; If so, confirm Let k be the updated lateral coordinates of the refraction points at each interface, and let k = k + 1; If not, control Decrease and return the proposed update values for each of the lateral coordinates in the (k+1)th iteration step, determined along the negative gradient direction. The steps.
7. The multi-layer dielectric structure full-focus imaging method according to any one of claims 2 to 6, characterized in that, The pixels are obtained by dividing the imaging area according to a grid. Each array element under the phased array probe is taken as the target array element, and each pixel is taken as the target pixel, including: Each element under the phased array probe is taken as a target element, and the pixels under the grid are traversed row by row for the target element to determine the corresponding target pixels. Based on the position coordinates of the target pixel in the coordinate system and a preset spatial coherence selection strategy, it is determined that the target pixel has related pixels corresponding to the target element, including: Based on the position coordinates of the target pixel in the coordinate system, determine whether the target pixel is the first pixel in the row where the target pixel is located; If it is not the first pixel, when it is determined that the left neighbor pixel of the target pixel in the row has the target horizontal coordinate corresponding to each interface refraction point, the left neighbor pixel is determined as the relevant pixel; when it is determined that the left neighbor pixel has not the target horizontal coordinate corresponding to each interface refraction point, the upper neighbor pixel of the target pixel in the column is determined as the relevant pixel.
8. The multi-layer dielectric structure full-focus imaging method as described in claim 7, characterized in that, When determining that the target pixel is the first pixel in the row containing the target pixel, the following steps are included: Determine the lateral coordinates of the intersection point of the target array element and the target pixel with the nth interface. for: Where 1≤n≤N-1, and N is the number of dielectric layers; Indicates the lateral coordinates of the target array element, ( , () represents the position coordinates of the target pixel. Represents the interface depth coordinates of the nth interface; Each of the above The initial values of the lateral coordinates of each interface refraction point along the path of the sound wave as it propagates from the target array element to the target pixel are used as the iterative initial values.
9. A multi-layer dielectric structure full-focus imaging system, characterized in that, include: A module is established to create a physical model and coordinate system for the M-layer dielectric structure and to determine the number of dielectric layers in the physical model for each pixel in the imaging area to be measured. M is an integer not less than 2; The sound propagation time determination module takes each array element under the phased array probe as the target array element and each pixel as the target pixel, and determines the sound propagation time relationship between the target array element and the target pixel based on the number of medium layers of the target pixel and the physical model. The judgment module is used to determine whether there are related pixels corresponding to the target element based on the position coordinates of the target pixel in the coordinate system and a preset spatial coherence selection strategy. If it exists, proceed to the initial value determination module; The initial value determination module is used to take the target position coordinates of each interface refraction point corresponding to the relevant pixel as the iterative initial value of the position coordinates of each interface refraction point that the sound wave passes through as it propagates from the target array element to the target pixel. The iteration module is used to iteratively determine the target sound propagation time and the target position coordinates of each interface refraction point when the sound propagates along the optimal propagation path between the target array element and the target pixel, based on the initial iteration value, the sound propagation time relationship and the preset gradient descent iteration strategy. The sampling and imaging module is used to sample the M-layer medium structure based on a full matrix data acquisition strategy to obtain full matrix data. Then, based on the target sound propagation time between each array element and each pixel and the full matrix data, the amplitude corresponding to each pixel is determined to obtain the full-focus imaging result of the imaging area.
10. A multi-layer dielectric structure full-focus imaging device, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the multilayer dielectric structure full-focus imaging method as described in any one of claims 1 to 8.